Skip to contents

Coming from Stata

Source: vignettes/articles/coming-from-stata.Rmd
coming-from-stata.Rmd

中文版

For chapter navigation, search and expandable alternatives, open the Stata to R Reference. The book has English and Chinese editions with a language switch on each chapter. This page remains available as the single-page reference.

If you know Stata, you already know what you want to do; the question is how to say it in R. This page lists Stata commands next to the R code that does the same job, using tidystata where it has a Stata-style function and the standard package everywhere else.

Look it up from the console

findit() searches the same list from R. Type one Stata word bare, or quote anything longer. Stata’s abbreviations work.

findit(reg)
#> # Linear regression [stats]
#> Stata: regress y x1 x2
#> R:     fit <- lm(y ~ x1 + x2, data = data)
#> # Note: summary(fit) shows the coefficient table; broom::tidy(fit) returns it as a data frame.
#> 
#> # 17 more examples. To see them all: findit(reg) |> print(n = Inf)
findit("merge 1:1 id, keep(match)")
#> # Keep matched rows only [dplyr]
#> Stata: merge 1:1 id using data2, keep(match)
#> R:     inner_join(data1, data2, by = "id")
#> # Note: dplyr retains overlapping non-key columns with suffixes instead of Stata master values, creates no _merge, and does not enforce a declared 1:1 key constraint.
#> 
#> # merge() [base]: Merge two data frames (base R)
#>   Related Stata: merge
#> 
#> # 17 more examples. To see them all: findit("merge 1:1 id, keep(match)") |> print(n = Inf)

You can also look up an R function (findit(left_join)) or ask a question (findit("how do I remove duplicate rows")). The console shows the best matches; the returned table contains all matches. Call findit() for the whole list. Descriptions follow R’s language setting; English and Chinese queries both work.

Comma-separated names customize compatible examples, such as findit(tab, sex, region). The default shows one preferred answer when the task is clear; findit(tab) |> print(n = Inf) shows all variants, selected alternative R approaches, qualifications and source links. Package-qualified lookups such as findit(dplyr::filter) distinguish functions with the same name.

This reference contains 514 command families, 181 Stata functions and 3 operators, with 1716 distinct Stata examples and 1782 paired R examples. Abbreviations and alternative R implementations do not increase the command count. The R index describes 949 function/package pairs.

Use the command tables

The tables cover tidystata and other R packages, including modelling, reshaping, survey analysis and graphics. For panel lags and leads, tidystata provides stata_lag(), stata_lead() and stata_diff(); specify the time column and use .by to identify panels.

The index is alphabetical; tables are grouped by subject and sorted by command, preferred answer and example. A package named beside a snippet is needed to run that snippet, but findit() itself does not load or install reference packages. The label names the primary package, rather than every helper dependency; the R function index identifies additional functions and their packages.

A note that begins Differs from Stata means the R code does the same job, but a default, the missing-value handling, an estimator or the output layout differs; the note says how. Other notes are practical tips. Most R snippets are run against controlled checks when the catalogue is built; this tests the R code itself, not numerical agreement with Stata. Missing-value comparisons, weights, factor reference levels, join rules and standard-error corrections need particular care. Known gaps are explained explicitly, and related search results are not a translation of unsupported syntax.

In the tables below, data is your dataset, data1 and data2 are the master and using data of a merge, and fit is a fitted model. Most entries use dplyr, so attach it with library(dplyr) or library(tidyverse).

Alphabetical command index

F · Fden · Ftail · about · ac · ado · adopath · alpha · ameans · anova · append · arch · areg · arfima · arfimasoc · args · arima · arimasoc · assert · assertnested · avplots · bayesboot · bayesmh · bayespredict · bayesreps · bayesstats · bayestest · betaden · betareg · binomial · binomialp · binomialtail · binormal · binreg · binscatter · biprobit · bitest · bitesti · bootstrap · boottest · boxcox · brier · browse · bsample · bsqreg · bstat · ca · camat · candisc · canon · capture · carryforward · cauchy · cauchyden · cauchytail · cc · cci · cd · ceil · cem · centile · cf · changeeol · char · checksum · chi2 · chi2den · chi2tail · ci · cii · ciwidth · clear · clock · cloglog · clonevar · cls · cluster · clustermat · cnsreg · codebook · coefplot · cofd · collapse · compare · compress · cond · confirm · constraint · continue · contract · contrast · copy · corr2data · correlate · corrgram · count · cpoisson · creturn · cross · cs · csdid · csi · cumsp · cumul · d. · dataex · datasignature · date · day · daysinmonth · decode · describe · destring · dfbeta · dfgls · dfuller · did_imputation · didregress · dir · discrim · display · distinct · do · dofc · dofm · dofq · dofy · dotplot · dow · doy · drawnorm · drdid · drop · ds · dstdize · dtable · duplicates · ebalance · egen · eivreg · elasticnet · encode · ereturn · error · esize · esizei · estat · estimates · eststo · esttab · etable · eventstudyinteract · exit · expand · expandcl · exponential · exponentialden · exponentialtail · export · f. · factor · factormat · fcast · fcollapse · file · filefilter · fillin · findfile · findit · firstdayofmonth · flevelsof · floor · fmerge · foreach · forecast · format · forvalues · fracreg · fralias · frame · fre · frget · frlink · frunalias · ftools · gammaden · gammap · gammaptail · gcollapse · gcontract · gdistinct · gduplicates · gegen · gencohort · generate · getmata · gettoken · gisid · glevelsof · glm · global · gquantiles · graph · greshape · gsort · gstats · gtoplevelsof · gunique · halfyear · hashsort · hausman · heatplot · heckman · help · hetregress · hexdump · hh · histogram · hotelling · hypergeometric · hypergeometricp · ibeta · ibetatail · icc · if · import · include · infile · infix · inlist · input · inrange · insobs · inspect · intreg · invF · invFtail · invbinomial · invbinomialtail · invcauchy · invcauchytail · invchi2 · invchi2tail · invexponential · invexponentialtail · invgammap · invgammaptail · invibeta · invibetatail · invlogistic · invlogistictail · invnF · invnFtail · invnbinomial · invnbinomialtail · invnchi2 · invnchi2tail · invnibeta · invnormal · invnt · invnttail · invpoisson · invpoissontail · invt · invttail · invtukeyprob · invweibull · invweibullph · invweibullphtail · invweibulltail · ipolate · iqreg · ir · irf · iri · isid · istdize · ivlpirf · ivreg2 · ivreghdfe · ivregress · jackknife · jdbc · joinby · kap · kdensity · keep · kmatch · ksmirnov · ktau · kwallis · l. · label · labelbook · labmask · lasso · levelsof · lincom · line · linktest · list · ln · lncauchyden · lnigammaden · lnmvnormalden · lnnormal · lnnormalden · local · log · logistic · logisticden · logistictail · logit · lookfor · lower · lowess · lpirf · lrtest · ltable · macro · manova · manovatest · margins · marginsplot · mark · markin · markout · marksample · matlist · matname · matrix · mca · mcc · mcci · mdesc · mds · mdsmat · mdy · mean · mecloglog · median · meglm · melogit · memory · menbreg · menl · meologit · meoprobit · mepoisson · meprobit · merge · meta · mhodds · mi · missing · missings · misstable · mixed · mkdir · mkmat · mlexp · mlogit · mm · mofd · month · mvdecode · mvencode · mvreg · mvtest · nF · nFden · nFtail · nbetaden · nbinomial · nbinomialp · nbinomialtail · nbreg · nchi2 · nchi2den · nchi2tail · nestreg · net · newey · nibeta · nl · nlcom · normal · normalden · notes · npnF · npnchi2 · npnt · nptrend · nt · nttail · numlabel · numlist · oaxaca · odbc · ologit · oneway · oprobit · order · orthog · orthpoly · outfile · outreg2 · pac · pca · pcamat · pchi · pcorr · pctile · pergram · permute · pnorm · poisson · poissonp · poissontail · post · postclose · postfile · power · pperron · ppmlhdfe · predict · predictnl · preserve · probit · program · proportion · prtest · prtesti · psdensity · psmatch2 · pstest · putdocx · putexcel · putmata · pwcompare · pwcorr · pwd · pwmean · qchi · qnorm · qofd · qqplot · qreg · quadchk · quantile · quarter · query · quietly · range · rangestat · ranksum · rbeta · rbinomial · rcauchy · rchi2 · rdbwdensity · rdbwselect · rddensity · rdrobust · real · recast · recode · regexm · regexr · regexreplaceall · regexs · reghdfe · regress · rename · replace · require · reshape · return · rexponential · rgamma · rhypergeometric · rlogistic · rnbinomial · rnormal · robvar · roccomp · roctab · rolling · rotate · rotatemat · round · rpoisson · rreg · rt · runiform · runtest · rvfplot · rweibull · sample · save · saveold · savespss · scalar · scatter · sdtest · sdtesti · search · sem · separate · seqvar · serrbar · set · setroot · signrank · signtest · simulate · sktest · sleep · slogit · snapshot · snapspan · sort · spbalance · spdistance · spearman · spgenerate · spikeplot · split · splitsample · spmatrix · spset · sqreg · ss · ssc · sspace · stack · stbase · stci · stcox · stcrreg · stdescribe · stem · stfill · stgen · stintreg · stir · stjoin · stmh · stptime · strate · streg · string · stritrim · strlen · strlower · strltrim · strmatch · strofreal · strpos · strproper · strreverse · strrtrim · strtoname · strtrim · strupper · sts · stset · stsplit · stsum · sttocc · stvary · subinstr · subinword · substr · sum · summarize · svmat · svy · svydescribe · svyset · swilk · symmetry · symmi · symplot · syntax · synth · sysdir · sysuse · t · tab1 · tab2 · tabdisp · tabi · table · tabodds · tabstat · tabulate · tden · tebalance · teffects · tempfile · tempname · tempvar · teoverlap · test · testnl · testparm · tetrachoric · timer · tnbreg · tobit · tokenize · tostring · tpoisson · truncreg · tsfill · tsfilter · tsline · tsreport · tsset · tssmooth · ttail · ttest · ttesti · tukeyprob · twoway · type · uchar · ucm · unab · unabcmd · unicode · univar · unzipfile · upper · use · uselabel · ustrlen · ustrlower · ustrnormalize · ustrpos · ustrregexm · ustrregexra · ustrregexrf · ustrreverse · ustrtrim · ustrupper · usubstr · var · vargranger · varstable · vec · vecrank · version · viewsource · vif · vioplot · webuse · weibull · weibullden · weibullph · weibullphden · weibullphtail · weibulltail · which · while · winsor2 · wntestq · word · wordcount · xcorr · xi · xpose · xtabond · xtcloglog · xtdescribe · xtdidregress · xtdpd · xtdpdsys · xtgee · xtgls · xthtaylor · xtile · xtivreg · xtline · xtlogit · xtologit · xtoprobit · xtpcse · xtpoisson · xtprobit · xtrc · xtreg · xtregar · xtscc · xtset · xtsum · xttest0 · xtunitroot · year · ym · yq · zinb · zip · zipfile · ztest · ztesti

Basics

Stata R Purpose, qualifications and sources 
cd "path/to/folder"
 
setwd("path/to/folder")
base 		Change the working directory
An RStudio project sets the working directory for you, so scripts can use relative paths.
Stata reference · R reference 
clear
 
rm(data)
base 		Remove a dataset from memory
Rarely needed: R holds many tables at once, each under its own name.
Stata reference · R reference 
dir
 
dir()
base 		List the files in the working directory
Stata reference · R reference 
display 2 + 2
 
2 + 2
base 		Calculate and show a result
Stata reference · R reference 
do analysis.do
 
source("analysis.R")
base 		Run a script file
Stata reference · R reference 
export delimited data.csv
 
readr::write_csv(data, "data.csv")
readr 		Export a CSV file
Stata reference · R reference 
export excel data.xlsx
 
writexl::write_xlsx(data, "data.xlsx")
writexl 		Export an Excel file
Stata reference · R reference 
findit command
 
findit(command)
tidystata 		Find the R equivalent of a Stata command
Bare words and comma-separated names work: findit(tab, sex, region). Add hints such as findit(tab, gen), or quote full Stata syntax. One clear example is shown by default; print(n = Inf) reveals every match. findit() with no argument returns the whole list.
Stata reference · R reference 
global path "path/to/folder"
 
path <- "path/to/folder"
base 		Store a value for later use
Any R object works like a macro: assign it once and reuse its name, as in setwd(path).
Stata reference · R reference 
help command
 
help(function_name)
utils 		Open the help page of a function
?function does the same. In RStudio you can also search the Help pane.
Stata reference · R reference 
import delimited data.csv
 
data <- readr::read_csv("data.csv")
readr 		Import a CSV file
RStudio's File > Import Dataset menu writes this code for you.
Stata reference · R reference 
import excel data.xlsx, firstrow
 
data <- readxl::read_excel("data.xlsx")
readxl 		Import an Excel sheet
Stata reference · R reference 
import sas data.sas7bdat
 
data <- haven::read_sas("data.sas7bdat")
haven 		Import a SAS file
Stata reference · R reference 
import spss data.sav
 
data <- haven::read_sav("data.sav")
haven 		Import an SPSS file
Stata reference · R reference 
log using results.log
 
sink("results.log")
base 		Record console output in a file
Most R users write an R Markdown or Quarto report instead of a log.
Stata reference · R reference 
net install package, from(url)
 
remotes::install_github("user/package")
remotes 		Install a package from GitHub
Stata reference · R reference 
pwd
 
getwd()
base 		Show the working directory
Stata reference · R reference 
save data.dta, replace
 
haven::write_dta(data, "data.dta")
haven 		Save as a Stata dataset
saveRDS(data, "data.rds") keeps every R type, such as factors and dates, exactly.
Stata reference · R reference 
saveold data.dta, version(13)
 
haven::write_dta(data, "data.dta", version = 13)
haven 		Save for an older Stata version
Stata reference · R reference 
search keyword
 
help.search("keyword")
utils 		Search all installed help pages
??keyword is a shortcut. To find the R version of a Stata command, use findit().
Stata reference · R reference 
ssc install package
 
install.packages("package")
utils 		Install a package
Unlike Stata, an installed package must be attached in every session with library(package).
Stata reference · R reference 
sysuse auto
 
data("mtcars")
datasets 		Load an example dataset
Stata reference · R reference 
use data.dta
 
data <- haven::read_dta("data.dta")
haven 		Open a Stata dataset
Variable and value labels are kept. Data saved in R's own format open with readRDS("data.rds") or load("data.RData").
Stata reference · R reference 
webuse nlswork
 
data <- haven::read_dta("https://www.stata-press.com/data/r18/nlswork.dta")
haven 		Load an example dataset from the Stata website
Stata reference · R reference

Explore data

Stata R Purpose, qualifications and sources 
anova y group
 
summary(aov(y ~ factor(group), data = data))
stats 		Analysis of variance
Stata reference · R reference 
browse
 
browse(data)
tidystata 		View the data in a spreadsheet window
Stata reference · R reference 
browse x1 x3-x5
 
browse(data, x1, x3:x5)
tidystata 		View selected variables
Ranges of variables use a colon (x3:x5), not a dash.
Stata reference · R reference 
browse *x
 
select(data, ends_with("x"))
dplyr 		Variables whose names end with given text
Stata reference · R reference 
browse *x*
 
select(data, contains("x"))
dplyr 		Variables whose names contain given text
Stata reference · R reference 
browse if x1 == 1 & x2 == 2
 
filter(data, x1 == 1 & x2 == 2)
dplyr 		Rows meeting a condition
Stata reference · R reference 
browse in -10/-1
 
tail(data, 10)
base 		The last rows
Stata reference · R reference 
browse in 1/10
 
head(data, 10)
base 		The first rows
Stata reference · R reference 
browse in 7/12
 
slice(data, 7:12)
dplyr 		Rows by position
Stata reference · R reference 
browse x*
 
select(data, starts_with("x"))
dplyr 		Variables whose names start with given text
Stata reference · R reference 
browse x* y*
 
browse(data, varlist("x* y*"))
tidystata 		View variables matching Stata wildcards
varlist() understands Stata's * and ? wildcards and works in any selection, such as select().
Stata reference · R reference 
ci mean x
 
t.test(data$x)$conf.int
stats 		Confidence interval of a mean
Add conf.level = 0.99 for ci mean x, level(99).
Stata reference · R reference 
codebook
 
codebook(data)
tidystata 		Describe every variable
Stata reference · R reference 
codebook, compact
 
summ(data)
tidystata 		One line of statistics per variable
Stata reference · R reference 
correlate x1 x2 x3
 
corr(data, x1, x2, x3)
tidystata 		Pearson correlations (complete cases)
Differs from Stata: Uses one common complete-case sample, like Stata correlate. Returns r and n in a tidy table; .wide = TRUE gives a lower-triangular table. Stata covariance, means and display-formatting options are not implemented.
Stata reference · R reference 
describe
 
variables(data)
tidystata 		List variables with their types and labels
glimpse(data) gives a compact view with the first values of each variable.
Stata reference · R reference 
describe using data.dta
 
haven::read_dta("data.dta", n_max = 0)
haven 		Variables in a file, without loading its data
Stata reference · R reference 
ds
 
ds(data)
tidystata 		List variable names
ds(data, .type = "table") returns the names as a one-column table; names(data) is base R.
Stata reference · R reference 
dtable x1 x2, by(group)
 
gtsummary::tbl_summary(data, by = group, include = c(x1, x2))
gtsummary 		Descriptive Table 1 (Stata 18+)
modelsummary::datasummary_skim(data) gives a quick overview without a by variable.
Stata reference · R reference 
factor x1 x2 x3
 
factanal(na.omit(select(data, x1, x2, x3)), factors = 1)
stats 		Factor analysis
Differs from Stata: factanal() uses maximum likelihood; psych::fa(..., fm = "pa") gives Stata's default principal-factor method.
Stata reference · R reference 
fre x
 
fre(data, x)
tidystata 		Frequency table with totals
Like Ben Jann's fre: tab() with a total row.
Stata reference · R reference 
inspect x
 
summ(data, x, .detail = TRUE)
tidystata 		Distribution, zeros and missing values of a variable
hist(data$x) adds the quick histogram that inspect draws.
Stata reference · R reference 
ksmirnov x, by(group)
 
ks.test(data$x[data$group == 0], data$x[data$group == 1])
stats 		Kolmogorov-Smirnov test of equal distributions
Differs from Stata: The R code assumes group codes 0 and 1; exact/asymptotic inference and tied-value handling can differ from Stata ksmirnov.
Stata reference · R reference 
ktau x y
 
cor.test(data$x, data$y, method = "kendall")
stats 		Kendall's tau
Differs from Stata: The rank coefficient has the same interpretation, but exact/asymptotic p-value selection and tie corrections may differ; match inference options separately.
Stata reference · R reference 
kwallis x, by(group)
 
kruskal.test(x ~ group, data = data)
stats 		Kruskal-Wallis test
Stata reference · R reference 
levelsof x
 
sort(unique(data$x))
base 		Distinct values of a variable
Differs from Stata: Store them for a loop: levels <- sort(unique(data$x)), then for (v in levels) { ... }. sort(unique()) excludes NA by default, but string ordering follows R collation and can differ from Stata sorting.
Stata reference · R reference 
list
 
print_all(data)
tidystata 		Print the first and last rows
Prints every row and column. Typing data prints the first rows; print_headtail(data) prints the first and last rows.
Stata reference · R reference 
list x
 
pull(data, x)
dplyr 		Print one variable
data$x does the same. select(data, x) keeps x as a one-column table.
Stata reference · R reference 
lookfor income
 
variables_search(data, "income")
tidystata 		Find variables by name or label
Stata reference · R reference 
mean x
 
summ(data, x, .stat = c("mean", "se"))
tidystata 		Mean with its standard error
Differs from Stata: t.test(data$x) also reports the confidence interval. The R snippet returns point summaries only (plus SE for mean); Stata also reports confidence intervals and estimation results.
Stata reference · R reference 
oneway y group
 
summary(aov(y ~ factor(group), data = data))
stats 		One-way analysis of variance
Differs from Stata: bartlett.test(y ~ group, data = data) gives the equal-variance test that oneway reports. aov() gives the ANOVA table but does not include the Bartlett equal-variance test printed by oneway; request bartlett.test() separately.
Stata reference · R reference 
pca x1 x2 x3
 
prcomp(na.omit(select(data, x1, x2, x3)), scale. = TRUE)
stats 		Principal components
Differs from Stata: Stata analyses the correlation matrix, so scale. = TRUE matches it. scale. = TRUE matches correlation PCA; signs of eigenvectors are arbitrary and the stored rotations and printed loadings have different normalization conventions.
Stata reference · R reference 
proportion x
 
prop.table(table(data$x))
base 		Proportion in each category
Differs from Stata: tab(data, x) shows the same percentages; prop.test() adds a confidence interval for one proportion. The R snippet returns point summaries only (plus SE for mean); Stata also reports confidence intervals and estimation results.
Stata reference · R reference 
prtest x == 0.5
 
prop.test(sum(data$x == 1, na.rm = TRUE), sum(!is.na(data$x)), p = 0.5, correct = FALSE)
stats 		Test of a proportion
Differs from Stata: correct = FALSE gives Stata's normal-approximation test without continuity correction. The uncorrected test statistic matches the normal approximation, but prop.test() confidence intervals use score inversion rather than the prtest Wald interval.
Stata reference · R reference 
pwcorr x1 x2 x3
 
pwcorr(data, x1, x2, x3)
tidystata 		Pairwise correlations
Differs from Stata: Gives r, n and p for each pair; .wide = TRUE gives the matrix. Base R: cor(select(data, x1, x2, x3), use = "pairwise.complete.obs"). pwcorr() returns pair counts and p-values in a tidy table; Stata prints a correlation matrix by default and needs options for additional reporting.
Stata reference · R reference 
ranksum x, by(group)
 
wilcox.test(x ~ group, data = data)
stats 		Wilcoxon rank-sum (Mann-Whitney) test
Differs from Stata: R can choose exact inference in small samples and applies continuity correction in its normal approximation; Stata defaults differ. Set exact and correct deliberately.
Stata reference · R reference 
signrank x = y
 
wilcox.test(data$x, data$y, paired = TRUE)
stats 		Wilcoxon signed-rank test
Differs from Stata: R can choose exact inference in small samples and applies continuity correction in its normal approximation; Stata defaults differ. Set exact and correct deliberately.
Stata reference · R reference 
sktest x
 
summ(data, x, .stat = c("skew", "kurtosis"))
tidystata 		Skewness and kurtosis
For a formal normality test use shapiro.test(data$x).
Stata reference · R reference 
spearman x y
 
cor.test(data$x, data$y, method = "spearman")
stats 		Spearman rank correlation
Differs from Stata: The rank coefficient has the same interpretation, but exact/asymptotic p-value selection and tie corrections may differ; match inference options separately.
Stata reference · R reference 
stem x
 
stem(data$x)
graphics 		Stem-and-leaf display
Stata reference · R reference 
summarize
 
summ(data)
tidystata 		Summary statistics for all variables
Differs from Stata: summary(data) is the base R version, with a less tidy layout. summ() also reports missing counts and has its own default statistic layout; select .stat explicitly when matching Stata output.
Stata reference · R reference 
summarize x
 
summ(data, x)
tidystata 		Summary statistics for one variable
Differs from Stata: summ() also reports missing counts and has its own default statistic layout; select .stat explicitly when matching Stata output.
Stata reference · R reference 
bysort group: summarize x
 
summ(data, x, .by = group)
tidystata 		Summary statistics by group
Differs from Stata: group_by(data, group) |> summ(x) gives the same result. summ() also reports missing counts and has its own default statistic layout; select .stat explicitly when matching Stata output.
Stata reference · R reference 
summarize x [aweight = w]
 
summ(data, x, .wt = w, .wt_type = "analytic")
tidystata 		Weighted summary statistics
Differs from Stata: Frequency weights are the default: summ(data, x, .wt = w). summ() also reports missing counts and has its own default statistic layout; select .stat explicitly when matching Stata output.
Stata reference · R reference 
summarize x, detail
 
summ(data, x, .detail = TRUE)
tidystata 		Detailed statistics, including percentiles
Differs from Stata: Pick single statistics with .stat, as in summ(data, x, .stat = "p25"). summ() also reports missing counts and has its own default statistic layout; select .stat explicitly when matching Stata output.
Stata reference · R reference 
swilk x
 
shapiro.test(data$x)
stats 		Shapiro-Wilk normality test
Differs from Stata: shapiro.test() accepts 3 to 5000 nonmissing observations; Stata swilk has different sample-size limits and reporting.
Stata reference · R reference 
tab1 x1 x2 x3
 
tab1(data, x1, x2, x3)
tidystata 		One frequency table per variable
Differs from Stata: tidystata retains missing categories by default; Stata tabulate excludes them unless missing is specified. Use .missing = FALSE to match Stata.
Stata reference · R reference 
tab2 x1 x2
 
tab2(data, x1, x2)
tidystata 		Cross-tabulation of two variables
Differs from Stata: With three or more variables, tab2(data, x1, x2, x3) returns every pairwise table, as Stata does. tidystata retains missing categories by default; Stata tabulate excludes them unless missing is specified. Use .missing = FALSE to match Stata.
Stata reference · R reference 
table x1 x2 x3
 
tab(data, x1, x2, x3)
tidystata 		Joint frequencies of several variables
Stata reference · R reference 
table x, statistic(mean y)
 
summ(data, y, .by = x, .stat = "mean")
tidystata 		A statistic by category (Stata 17+ table)
Stata reference · R reference 
tabstat x1 x2, by(group) stat(mean sd)
 
tabstat(data, x1, x2, .by = group, .stat = c("mean", "sd"))
tidystata 		Chosen statistics by group
Differs from Stata: R retains missing grouping keys; Stata tabstat by() and tabulate summarize() exclude missing grouping categories unless requested. Check the intended grouping sample.
Stata reference · R reference 
tabulate x
 
tab(data, x)
tidystata 		One-way frequency table
Differs from Stata: tidystata retains missing categories by default; Stata tabulate excludes them unless missing is specified. Use .missing = FALSE to match Stata.
Stata reference · R reference 
tabulate x1 x2
 
tab2(data, x1, x2)
tidystata 		Two-way cross-tabulation
Differs from Stata: fre2(data, x1, x2) adds row and column totals. tidystata retains missing categories by default; Stata tabulate excludes them unless missing is specified. Use .missing = FALSE to match Stata.
Stata reference · R reference 
bysort g: tabulate x
 
tab(data, x, .by = g)
tidystata 		Frequency table by group
Differs from Stata: tidystata retains missing categories by default; Stata tabulate excludes them unless missing is specified. Use .missing = FALSE to match Stata.
Stata reference · R reference 
tabulate x [aweight = w]
 
tab(data, x, .wt = w, .wt_type = "analytic")
tidystata 		Frequencies with analytic (sampling) weights
Differs from Stata: tidystata retains missing categories by default; Stata tabulate excludes them unless missing is specified. Use .missing = FALSE to match Stata.
Stata reference · R reference 
tabulate x [fweight = n]
 
tab(data, x, .wt = n)
tidystata 		Frequencies with frequency weights
Differs from Stata: tidystata retains missing categories by default; Stata tabulate excludes them unless missing is specified. Use .missing = FALSE to match Stata.
Stata reference · R reference 
tabulate x, missing
 
tab(data, x)
tidystata 		Frequencies including missing values
tab() always lists missing values as a separate row.
Stata reference · R reference 
tabulate x, summarize(y)
 
summ(data, y, .by = x)
tidystata 		Summary of y within each category of x
Differs from Stata: R retains missing grouping keys; Stata tabstat by() and tabulate summarize() exclude missing grouping categories unless requested. Check the intended grouping sample.
Stata reference · R reference 
tabulate x1 x2, chi2
 
tab2(dplyr::filter(data, !is.na(x1), !is.na(x2)), x1, x2, .chi2 = TRUE)
tidystata 		Chi-square test of independence
Differs from Stata: Filter missing values on both variables for plain Stata tabulate, chi2. tab2 otherwise includes its displayed missing categories in the Pearson test. tidystata retains missing categories by default; Stata tabulate excludes them unless missing is specified. Use .missing = FALSE to match Stata.
Stata reference · R reference 
tabulate x1 x2, exact
 
fisher.test(data$x1, data$x2)
stats 		Fisher's exact test
Stata reference · R reference 
ttest x, by(group)
 
t.test(x ~ group, data = data)
stats 		Two-sample t test
Differs from Stata: R's default is Welch's test; add var.equal = TRUE for Stata's default equal-variance test.
Stata reference · R reference 
univar x1 x2
 
summ(data, x1, x2, .detail = TRUE)
tidystata 		Detailed univariate statistics
Stata reference · R reference

Check data

Stata R Purpose, qualifications and sources 
assert x > 0
 
assert(data, x > 0)
tidystata 		Stop when a condition fails
Differs from Stata: Reports the number of contradictions and returns the data, so it fits in a pipe. Missing x counts as a contradiction, unlike Stata, where missing is larger than any number: write assert(data, is.na(x) | x > 0) for Stata's behaviour.
Stata reference · R reference 
confirm variable x
 
stopifnot("x" %in% names(data))
base 		Check that a variable exists
Stata reference · R reference 
count
 
nrow(data)
base 		Number of rows
Stata reference · R reference 
count if x > 0
 
sum(data$x > 0, na.rm = TRUE)
base 		Count rows meeting a condition
Differs from Stata: Stata numeric missing values compare larger than every finite number; R comparisons return NA. Handle missing values explicitly before applying the condition.
Stata reference · R reference 
distinct x1 x2, joint
 
nrow(distinct(data, x1, x2))
dplyr 		Number of distinct combinations
Stata reference · R reference 
duplicates list x1 x2
 
duplicates_list(data, x1, x2)
tidystata 		List duplicated rows
Stata reference · R reference 
duplicates report x1 x2
 
duplicates_report(data, x1, x2)
tidystata 		Count duplicated rows
Stata reference · R reference 
duplicates tag x1 x2, generate(dup)
 
duplicates_tag(data, x1, x2, .name = "dup")
tidystata 		Mark duplicated rows
Stata reference · R reference 
isid id wave
 
isid(data, id, wave)
tidystata 		Check that variables identify the rows
Stata reference · R reference 
mdesc
 
summ(data, .stat = "missing")
tidystata 		Missing values per variable
Stata reference · R reference 
misstable patterns
 
misstable(data, .patterns = TRUE)
tidystata 		Missing-value patterns
1 marks an observed value and 0 a missing one, as in Stata.
Stata reference · R reference 
misstable summarize
 
misstable(data)
tidystata 		Missing values per variable
Base R: colSums(is.na(data)).
Stata reference · R reference

Labels

Stata R Purpose, qualifications and sources 
decode x, generate(str_x)
 
mutate(data, str_x = decode(x))
tidystata 		Turn labelled values into text
Stata reference · R reference 
encode str_x, generate(x)
 
mutate(data, x = factor(str_x))
dplyr 		Turn text into labelled categories
Differs from Stata: factor() creates integer level indices with an R class; encode creates numeric codes with a separate value-label definition. Collation can change code ordering.
Stata reference · R reference 
label define gender 1 "male" 2 "female"
 
mutate(data, gender = factor(gender, levels = 1:2, labels = c("male", "female")))
dplyr 		Label every value of a categorical variable
In R a labelled categorical variable is usually a factor. Give levels explicitly so each code gets the right label.
Stata reference · R reference 
label drop yesno
 
mutate(data, x = haven::zap_labels(x))
haven 		Remove value labels
Stata reference · R reference 
label list x
 
levels(data$x)
base 		Show the categories of a variable
For labelled data read with haven, use haven::print_labels(data$x).
Stata reference · R reference 
label values x yesno
 
mutate(data, x = haven::labelled(x, yesno))
haven 		Attach a value label to a variable
Define the labels once, as yesno <- c(No = 0, Yes = 1), then attach them to any variable; factor() is the usual R alternative.
Stata reference · R reference 
label variable x "Variable name"
 
labelled::set_variable_labels(data, x = "Variable name")
labelled 		Label a variable
Labels read from Stata files are kept; variables() and codebook() display them.
Stata reference · R reference 
label define x 9 "Unknown" 10 "Refused"
 
mutate(data, x = haven::labelled(x, c(Unknown = 9, Refused = 10)))
haven 		Label some values only
Partial value labels are rare in R; haven keeps them so they survive a round trip to Stata. tab() and codebook() show them.
Stata reference · R reference 
notes x: Collected in wave 2
 
attr(data$x, "note") <- "Collected in wave 2"
base 		Attach a note to a variable
R has no notes system; an attribute or a comment in the script is the usual substitute.
Stata reference · R reference

Construct and modify variables

Stata R Purpose, qualifications and sources 
generate y = ceil(x)
 
mutate(data, y = ceiling(x))
base 		Round up
Stata reference · R reference 
clonevar x2 = x
 
mutate(data, x2 = x)
dplyr 		Copy a variable
Differs from Stata: mutate(x2 = x) copies the R vector and attributes, but does not reproduce all Stata clonevar characteristics and notes.
Stata reference · R reference 
collapse (mean) x1 x2, by(group)
 
summarise(data, across(c(x1, x2), ~ mean(.x, na.rm = TRUE)), .by = group)
dplyr 		Replace the data with group means
Differs from Stata: Assign the result to replace the data, as collapse does: data <- summarise(...). summ(data, x1, x2, .by = group) shows a summary table instead. R retains missing grouping keys; Stata tabstat by() and tabulate summarize() exclude missing grouping categories unless requested. Check the intended grouping sample.
Stata reference · R reference 
collapse (count) n = x (sum) total = y, by(group)
 
summarise(data, n = sum(!is.na(x)), total = sum(y, na.rm = TRUE), .by = group)
dplyr 		Different statistics for different variables
Stata reference · R reference 
generate y = cond(x > 0, 1, 0)
 
mutate(data, y = if_else(x > 0, 1, 0))
dplyr 		One value if a condition holds, another if not
Differs from Stata: With missing x, Stata's cond() takes the first value (missing counts as large); if_else() returns NA. Stata numeric missing values compare larger than every finite number; R comparisons return NA. Handle missing values explicitly before applying the condition.
Stata reference · R reference 
destring str_x, generate(num_x)
 
mutate(data, num_x = destring(str_x))
tidystata 		Text to numbers
Differs from Stata: Text that is not a number becomes NA with a warning, as with force. destring(str_x, ignore = "kg") removes those characters first, like ignore(), so "10 kg" gives 10. Invalid numeric text becomes NA with a warning here; Stata destring without force refuses conversion of a variable containing such text.
Stata reference · R reference 
egen bins = cut(x), group(10)
 
mutate(data, bins = xtile(x, 10))
tidystata 		Ten groups of about equal size
Differs from Stata: egen cut(), group() has its own grouping rule; xtile() follows Stata's xtile, the usual command for quantile groups.
Stata reference · R reference 
egen first = tag(group)
 
mutate(data, first = is_first(group))
tidystata 		Mark the first row of each group
Differs from Stata: Gives TRUE/FALSE; rows with a missing group are FALSE, as in Stata, unless .missing = TRUE. The R result is logical TRUE/FALSE rather than numeric 1/0; use as.integer() when the stored type matters.
Stata reference · R reference 
egen id = group(g1 g2)
 
data |> group_by(g1, g2) |> mutate(id = cur_group_id()) |> ungroup()
dplyr 		Number the groups
Differs from Stata: cur_group_id() also numbers groups with missing keys; egen group() excludes those rows without the missing option. String collation can change ID ordering.
Stata reference · R reference 
egen x_avg = mean(x)
 
mutate(data, x_avg = mean(x, na.rm = TRUE))
dplyr 		Sample statistic as a new variable
Differs from Stata: R functions keep missing values unless na.rm = TRUE; Stata's egen ignores them. For an all-missing group mean(..., na.rm = TRUE) gives NaN; Stata egen mean() gives numeric missing. Normalize NaN if that distinction matters.
Stata reference · R reference 
egen x_sum = rowtotal(x1-x5)
 
mutate(data, x_sum = row_sum(x1:x5))
tidystata 		Row total
Stata reference · R reference 
bysort group: egen med = median(x)
 
mutate(data, med = median(x, na.rm = TRUE), .by = group)
dplyr 		Group median
Stata reference · R reference 
bysort group: egen mn = min(x)
 
dplyr::mutate(data, mn = if (all(is.na(x))) NA_real_ else min(x, na.rm = TRUE), .by = group)
dplyr 		Group minimum
Differs from Stata: All-missing groups return NA without evaluating min() on an empty set. All-missing groups produce -Inf/Inf and a warning in R max/min with na.rm = TRUE; Stata egen max()/min() returns missing.
Stata reference · R reference 
bysort group: egen mx = max(x)
 
dplyr::mutate(data, mx = if (all(is.na(x))) NA_real_ else max(x, na.rm = TRUE), .by = group)
dplyr 		Group maximum
Differs from Stata: All-missing groups return NA without evaluating max() on an empty set. All-missing groups produce -Inf/Inf and a warning in R max/min with na.rm = TRUE; Stata egen max()/min() returns missing.
Stata reference · R reference 
bysort group: egen n = count(x)
 
mutate(data, n = sum(!is.na(x)), .by = group)
dplyr 		Number of observed values in each group
Stata reference · R reference 
bysort group: egen s = sd(x)
 
mutate(data, s = sd(x, na.rm = TRUE), .by = group)
dplyr 		Group standard deviation
Stata reference · R reference 
bysort group: egen tot = total(x)
 
mutate(data, tot = sum(x, na.rm = TRUE), .by = group)
dplyr 		Group total
Stata reference · R reference 
bysort group: egen x_avg = mean(x)
 
mutate(data, x_avg = mean(x, na.rm = TRUE), .by = group)
dplyr 		Group statistic as a new variable
Differs from Stata: For an all-missing group mean(..., na.rm = TRUE) gives NaN; Stata egen mean() gives numeric missing. Normalize NaN if that distinction matters.
Stata reference · R reference 
egen bins = cut(x), at(0(10)100)
 
mutate(data, bins = cut(x, breaks = seq(0, 100, 10), right = FALSE))
base 		Groups by given cutpoints
Differs from Stata: right = FALSE makes each interval include its lower cutpoint, as egen cut() does. cut_length(x, 10) instead splits the range into 10 equal widths. cut() returns a factor of interval labels; egen cut() stores numeric lower endpoints even when interval closure matches.
Stata reference · R reference 
egen first = rowfirst(x1-x5)
 
mutate(data, first = coalesce(x1, x2, x3, x4, x5))
dplyr 		First observed value across variables
Stata reference · R reference 
egen id = concat(a b), punct(_)
 
mutate(data, id = paste(a, b, sep = "_"))
base 		Join variables into one string
Differs from Stata: paste() turns NA into the text "NA"; egen concat() follows Stata missing-value and numeric-display conventions.
Stata reference · R reference 
egen last = rowlast(x1-x5)
 
mutate(data, last = coalesce(x5, x4, x3, x2, x1))
dplyr 		Last observed value across variables
Stata reference · R reference 
egen n9 = anycount(x1-x5), values(9)
 
mutate(data, n9 = rowSums(across(x1:x5, ~ .x %in% 9)))
dplyr 		Count how many variables equal a value
Stata reference · R reference 
egen p = pc(x)
 
mutate(data, p = pc(x))
tidystata 		Percentage of the total
pc(x, prop = TRUE) gives proportions, like the prop option.
Stata reference · R reference 
egen r = rank(x)
 
mutate(data, r = rank(x, na.last = "keep"))
base 		Rank, with ties averaged
rank() gives tied values their average rank, like egen rank(); dplyr::min_rank() gives them the lowest.
Stata reference · R reference 
egen s = seq(), from(1) to(3)
 
mutate(data, s = rep_len(1:3, n()))
dplyr 		Repeating sequence
Stata reference · R reference 
egen x_avg = rowmean(x1-x5)
 
mutate(data, x_avg = row_mean(x1:x5))
tidystata 		Row mean
Stata reference · R reference 
egen x_max = rowmax(x1-x5)
 
mutate(data, x_max = row_max(x1:x5))
tidystata 		Row maximum
Stata reference · R reference 
egen x_med = rowmedian(x1-x5)
 
mutate(data, x_med = row_median(x1:x5))
tidystata 		Row median
Stata reference · R reference 
egen x_mi = rowmiss(x1-x5)
 
mutate(data, x_mi = row_miss(x1:x5))
tidystata 		Number of missing values in each row
Stata reference · R reference 
egen x_min = rowmin(x1-x5)
 
mutate(data, x_min = row_min(x1:x5))
tidystata 		Row minimum
Stata reference · R reference 
egen x_n = rownonmiss(x1-x5)
 
mutate(data, x_n = row_non_miss(x1:x5))
tidystata 		Number of observed values in each row
Stata reference · R reference 
egen x_nv = rownvals(x1-x5)
 
mutate(data, x_nv = row_nvals(x1:x5))
tidystata 		Number of distinct values in each row
From egenmore (SSC). Missing values are ignored unless .missing = TRUE, like the missing option.
Stata reference · R reference 
egen x_sd = rowsd(x1-x5)
 
mutate(data, x_sd = row_sd(x1:x5))
tidystata 		Row standard deviation
Stata reference · R reference 
egen z = std(x)
 
mutate(data, z = std(x))
tidystata 		Standardise a variable
Base R: as.vector(scale(x)). Add .by = g for by g: egen.
Stata reference · R reference 
generate y = floor(x)
 
mutate(data, y = floor(x))
base 		Round down
Stata reference · R reference 
format x %9.2f
 
mutate(data, x_text = sprintf("%.2f", x))
base 		Display numbers with two decimals
R stores the number unchanged and formats it only when printing, for example with round(x, 2) or sprintf().
Stata reference · R reference 
generate x2 = x^2
 
mutate(data, x2 = x^2)
dplyr 		Create a variable
Assign the result to keep it: data <- mutate(data, ...). R has no separate generate and replace.
Stata reference · R reference 
generate N = _N
 
mutate(data, N = n())
dplyr 		Number of rows
Add .by = group for the number of rows in each group, like bysort group: generate N = _N.
Stata reference · R reference 
generate n = _n
 
mutate(data, n = row_number())
dplyr 		Row number
Stata reference · R reference 
generate days = d2 - d1
 
mutate(data, days = as.numeric(d2 - d1))
base 		Days between two dates
Stata reference · R reference 
generate dummy = (x1 == 1 & x2 == 2)
 
dplyr::mutate(data, dummy = as.integer(x1 %in% 1 & x2 %in% 2))
dplyr 		Dummy variable from a condition
Differs from Stata: Stata numeric equality is false for missing operands; %in% keeps the dummy zero when either input is missing. The R result is logical TRUE/FALSE rather than numeric 1/0; use as.integer() when the stored type matters.
Stata reference · R reference 
generate lag_x = x[_n-1]
 
mutate(data, lag_x = lag(x))
dplyr 		Value from the previous row
This shifts rows, like x[_n-1], and ignores gaps in time. See L. for panel lags.
Stata reference · R reference 
generate lead_x = x[_n+2]
 
mutate(data, lead_x = lead(x, 2))
dplyr 		Value from two rows ahead
Stata reference · R reference 
generate x2 = x^2 if x > 0
 
mutate(data, x2 = if_else(x > 0, x^2, NA))
dplyr 		Create a variable under a condition
In Stata missing is larger than any number, so x > 0 is true for missing x. In R the condition is NA and so is the result.
Stata reference · R reference 
generate x2 = x^2 in 11/20
 
mutate(data, x2 = if_else(row_number() %in% 11:20, x^2, NA))
dplyr 		Create a variable for some rows
Stata reference · R reference 
generate y = inlist(x, 1, 3, 5)
 
mutate(data, y = x %in% c(1, 3, 5))
base 		Whether x equals any listed value
Differs from Stata: %in% gives FALSE for missing x, as inlist() does. R returns a logical indicator rather than the numeric 1/0 stored by Stata generate; use as.integer() if the stored type matters.
Stata reference · R reference 
generate y = inrange(x, 1, 5)
 
mutate(data, y = between(x, 1, 5))
dplyr 		Whether x lies within a range
Differs from Stata: between() returns NA for missing x, whereas Stata's inrange() returns 0.
Stata reference · R reference 
generate y = ln(x)
 
mutate(data, y = log(x))
base 		Natural logarithm
Differs from Stata: log10() matches Stata's log10(). R's log(0) is -Inf rather than missing. R log(0) gives -Inf and a negative input gives NaN with a warning; Stata ln() gives numeric missing for these invalid inputs.
Stata reference · R reference 
generate m = missing(x)
 
mutate(data, m = is.na(x))
base 		Whether a value is missing
Differs from Stata: R returns a logical indicator rather than the numeric 1/0 stored by Stata generate; use as.integer() if the stored type matters.
Stata reference · R reference 
mvdecode x, mv(-9)
 
mutate(data, x = mvdecode(x, -9))
tidystata 		Turn missing-value codes into missing
mvdecode(x, c(-8, -9)) handles several codes, like mv(-8 -9).
Stata reference · R reference 
mvencode x, mv(-9)
 
mutate(data, x = mvencode(x, -9))
tidystata 		Turn missing values into a code
Stata reference · R reference 
pctile p = x, nquantiles(4)
 
summ(data, x, .stat = c("p25", "median", "p75"))
tidystata 		Quantile cutpoints
Differs from Stata: summ() uses Stata's percentile rule. quantile(x, type = 2) is close, but treats some whole positions, such as 100 * 0.3, as fractional. summ() returns percentile values in a summary table; pctile creates a dataset column containing the cutpoints.
Stata reference · R reference 
recast double x
 
mutate(data, x = as.double(x))
base 		Change the storage type
R chooses storage automatically, so there is no compress either.
Stata reference · R reference 
recode x (1/5 = 1) (6/10 = 2) (11/15 = 3) (else = .), generate(x3g)
 
mutate(data, x3g = case_when(between(x, 1, 5) ~ 1, between(x, 6, 10) ~ 2, between(x, 11, 15) ~ 3))
dplyr 		Recode ranges of values
Values that match no condition become NA, like else = . in Stata.
Stata reference · R reference 
recode x (9 = .)
 
mutate(data, x = na_if(x, 9))
dplyr 		Turn one code into missing
Stata reference · R reference 
rename x x2
 
rename(data, x2 = x)
dplyr 		Rename a variable
Note the order: new name = old name.
Stata reference · R reference 
rename (a b) (c d)
 
rename(data, c = a, d = b)
dplyr 		Rename several variables
Stata reference · R reference 
rename * *_old
 
rename_with(data, ~ paste0(.x, "_old"))
dplyr 		Add a suffix to every name
Stata reference · R reference 
rename *, lower
 
rename_with(data, tolower)
dplyr 		Make all names lower case
Stata reference · R reference 
rename x*, upper
 
rename_with(data, toupper, starts_with("x"))
dplyr 		Upper-case some names
Stata reference · R reference 
replace x = x^2
 
mutate(data, x = x^2)
dplyr 		Change a variable
Stata reference · R reference 
replace x = x^2 if x > 0
 
mutate(data, x = if_else(x > 0, x^2, x))
dplyr 		Change a variable under a condition
Differs from Stata: Stata numeric missing values compare larger than every finite number; R comparisons return NA. Handle missing values explicitly before applying the condition.
Stata reference · R reference 
replace x = x^2 in 11/20
 
mutate(data, x = if_else(row_number() %in% 11:20, x^2, x))
dplyr 		Change a variable for some rows
Stata reference · R reference 
generate z = rnormal()
 
mutate(data, z = rnorm(n()))
stats 		Normal random numbers
Differs from Stata: R and Stata use different random-number streams; the same numeric seed does not reproduce the same draws or selected rows.
Stata reference · R reference 
generate y = round(x, 0.1)
 
mutate(data, y = round(x, 1))
base 		Round
Differs from Stata: R's second argument is the number of decimals. R rounds halves to even (round(0.5) is 0); Stata rounds them up.
Stata reference · R reference 
generate u = runiform()
 
mutate(data, u = runif(n()))
stats 		Uniform random numbers
Differs from Stata: R and Stata use different random-number streams; the same numeric seed does not reproduce the same draws or selected rows.
Stata reference · R reference 
generate cs = sum(x)
 
mutate(data, cs = cumsum(coalesce(x, 0)))
dplyr 		Running sum
Stata's sum() function treats missing values as 0, hence coalesce(x, 0).
Stata reference · R reference 
tabulate x, generate(x_)
 
as_dummy(data, x)
tidystata 		Dummy variables for every category
Differs from Stata: as_dummy(data, x, .names = "index") numbers the columns x_1, x_2, ... as Stata does. Also: fastDummies::dummy_cols(data, "x"). as_dummy() defaults to value-based column names; .names = "index" gives numbered Stata-style suffixes. Check the missing-category indicator policy too.
Stata reference · R reference 
tostring num_x, generate(str_x)
 
mutate(data, str_x = tostring(num_x))
tidystata 		Numbers to text
Differs from Stata: Labelled values keep their codes; decode() gives the value labels instead. Formatting and retained precision can differ from Stata tostring; choose an explicit format when exact text or round-trip precision matters.
Stata reference · R reference 
winsor2 x, cuts(1 99)
 
mutate(data, x_w = datawizard::winsorize(x, threshold = 0.01))
datawizard 		Winsorise at the 1st and 99th percentiles
Differs from Stata: datawizard uses R quantile cutpoints while winsor2 uses Stata percentile rules; small samples and ties can therefore yield different caps.
Stata reference · R reference 
xtile q = x, nquantiles(4)
 
mutate(data, q = xtile(x, 4))
tidystata 		Quantile groups
xtile() follows Stata's rule, so equal values stay in the same group.
Stata reference · R reference 
xtile q = x [aweight = w], nquantiles(4)
 
mutate(data, q = xtile(x, 4, wt = w))
tidystata 		Weighted quantile groups
Stata reference · R reference

Select variables and rows

Stata R Purpose, qualifications and sources 
drop if x == 1
 
filter(data, x != 1 | is.na(x))
dplyr 		Drop rows meeting a condition
Stata keeps rows with missing x here; | is.na(x) does the same in R.
Stata reference · R reference 
drop in 1/4
 
slice(data, -(1:4))
dplyr 		Drop rows by position
Stata reference · R reference 
drop x
 
select(data, -x)
dplyr 		Drop a variable
Stata reference · R reference 
drop if missing(x)
 
filter(data, !is.na(x))
dplyr 		Drop rows where x is missing
Stata reference · R reference 
drop x1 x3-x5
 
select(data, -c(x1, x3:x5))
dplyr 		Drop several variables
Positions work too: select(data, -c(1, 3:5)).
Stata reference · R reference 
duplicates drop
 
duplicates_drop(data)
tidystata 		Drop rows that duplicate an earlier row
Stata reference · R reference 
duplicates drop x1 x2, force
 
duplicates_drop(data, x1, x2, .force = TRUE)
tidystata 		Keep the first row for each key
distinct(data, x1, x2, .keep_all = TRUE) is the dplyr version.
Stata reference · R reference 
keep if x1 == 1 & x2 == 2
 
filter(data, x1 == 1 & x2 == 2)
dplyr 		Keep rows meeting a condition
filter() drops rows where the condition is NA, so rows with missing values never pass.
Stata reference · R reference 
keep in 1/10
 
head(data, 10)
base 		Keep the first rows
Stata reference · R reference 
keep x1 x3-x5
 
select(data, x1, x3:x5)
dplyr 		Keep some variables
Numbers select by position: select(data, 1, 3:5). select_or_transmute() also computes new columns.
Stata reference · R reference 
keep if !missing(x)
 
filter(data, !is.na(x))
dplyr 		Keep rows where x is observed
Stata reference · R reference 
keep if inlist(x, 1, 3, 5)
 
filter(data, x %in% c(1, 3, 5))
dplyr 		Keep rows with listed values
Stata reference · R reference 
keep if inrange(x, 5000, 10000)
 
filter(data, between(x, 5000, 10000))
dplyr 		Keep rows within a range
Stata reference · R reference 
keep in -10/-1
 
tail(data, 10)
base 		Keep the last rows
Stata reference · R reference 
keep in 7/12
 
slice(data, 7:12)
dplyr 		Keep rows by position
Stata reference · R reference 
keep x*
 
select(data, varlist("x*"))
tidystata 		Keep variables matching Stata wildcards
Stata reference · R reference 
set seed 2024; sample 10
 
data |> slice_sample(prop = 0.1) |> set_seed(2024)
tidystata 		Reproducible random 10% sample
Differs from Stata: R and Stata use different random-number streams; the same numeric seed does not reproduce the same draws or selected rows.
Stata reference · R reference 
set seed 2024; sample 1000, count
 
data |> slice_sample(n = 1000) |> set_seed(2024)
tidystata 		Reproducible random sample of 1000 rows
Differs from Stata: set_seed() sets the seed before the sampling step runs, so the pipe stays in one line. R and Stata use different random-number streams; the same numeric seed does not reproduce the same draws or selected rows.
Stata reference · R reference

Sort

Stata R Purpose, qualifications and sources 
gsort x1 -x2
 
arrange(data, x1, desc(x2))
dplyr 		Sort ascending and descending
Stata reference · R reference 
order x1-x4
 
relocate(data, x1:x4)
dplyr 		Move variables to the front
Stata reference · R reference 
order x1-x4, before(x5)
 
relocate(data, x1:x4, .before = x5)
dplyr 		Move variables before another
Stata reference · R reference 
sort x1 x2
 
arrange(data, x1, x2)
dplyr 		Sort in ascending order
Stata reference · R reference

Combine data

Stata R Purpose, qualifications and sources 
append using data2
 
bind_rows(data1, data2)
dplyr 		Stack two datasets
Differs from Stata: bind_rows() has R type-promotion and attribute rules; .id creates string source identifiers rather than the numeric codes of append, generate().
Stata reference · R reference 
append using data2, generate(source)
 
dplyr::bind_rows(data1, data2, .id = "source") |> dplyr::mutate(source = as.integer(source) - 1L)
dplyr 		Stack datasets and record where each row came from
Differs from Stata: Uses numeric source IDs: zero for the master and one for the first using dataset, as Stata append, generate() does. bind_rows() has R type-promotion and attribute rules; .id creates string source identifiers rather than the numeric codes of append, generate().
Stata reference · R reference 
cross using data2
 
cross_join(data1, data2)
dplyr 		Every combination of rows from two datasets
Differs from Stata: cross_join() suffixes overlapping column names; Stata cross keeps master values for overlapping names. Rename overlaps first to obtain the same columns.
Stata reference · R reference 
joinby id using data2
 
inner_join(data1, data2, by = "id", relationship = "many-to-many")
dplyr 		All pairs of rows within each id
Differs from Stata: dplyr retains overlapping non-key columns with suffixes instead of Stata master values, creates no _merge, and does not enforce a declared 1:1 key constraint.
Stata reference · R reference 
merge 1:1 _n using data2
 
bind_cols(data1, data2)
dplyr 		Put two datasets side by side by row position
Differs from Stata: bind_cols() needs matching row counts (or recyclable one-row inputs) and repairs duplicate names; merge 1:1 _n can retain unmatched rows.
Stata reference · R reference 
merge 1:1 id using data2
 
stata_merge(data1, data2, "1:1", by = "id")
tidystata 		Merge with a match report
Like Stata, all rows are kept and a .merge column records the match result. full_join(data1, data2, by = "id") merges without the report.
Stata reference · R reference 
merge 1:m id using data2
 
stata_merge(data1, data2, "1:m", by = "id")
tidystata 		One-to-many merge with a match report
Stata reference · R reference 
merge m:1 id using data2
 
stata_merge(data1, data2, "m:1", by = "id")
tidystata 		Many-to-one merge with a match report
Stata reference · R reference 
merge 1:1 id using data2, keep(master match)
 
left_join(data1, data2, by = "id")
dplyr 		Keep all rows of the master data
Differs from Stata: dplyr retains overlapping non-key columns with suffixes instead of Stata master values, creates no _merge, and does not enforce a declared 1:1 key constraint.
Stata reference · R reference 
merge 1:1 id using data2, keep(master) keepusing(id) nogenerate
 
anti_join(data1, data2, by = "id")
dplyr 		Keep rows of data1 whose id is not in data2
Differs from Stata: semi_join()/anti_join() select rows by key membership but do not enforce the unique keys required by Stata merge 1:1; validate uniqueness separately.
Stata reference · R reference 
merge 1:1 id using data2, keep(match)
 
inner_join(data1, data2, by = "id")
dplyr 		Keep matched rows only
Differs from Stata: dplyr retains overlapping non-key columns with suffixes instead of Stata master values, creates no _merge, and does not enforce a declared 1:1 key constraint.
Stata reference · R reference 
merge 1:1 id using data2, keep(match) keepusing(id) nogenerate
 
semi_join(data1, data2, by = "id")
dplyr 		Keep rows of data1 whose id is in data2
Differs from Stata: semi_join()/anti_join() select rows by key membership but do not enforce the unique keys required by Stata merge 1:1; validate uniqueness separately.
Stata reference · R reference 
merge 1:1 id using data2, keep(using match)
 
right_join(data1, data2, by = "id")
dplyr 		Keep all rows of the using data
Differs from Stata: dplyr retains overlapping non-key columns with suffixes instead of Stata master values, creates no _merge, and does not enforce a declared 1:1 key constraint.
Stata reference · R reference 
merge 1:1 id1 id2 using data2, nogenerate
 
full_join(data1, data2, by = c("id1", "id2"))
dplyr 		Merge on two keys
Differs from Stata: Variables with the same name in both tables get suffixes; choose them with suffix = c("_data1", "_data2"). dplyr retains overlapping non-key columns with suffixes instead of Stata master values, creates no _merge, and does not enforce a declared 1:1 key constraint.
Stata reference · R reference

Reshape

Stata R Purpose, qualifications and sources 
reshape long var, i(id) j(year)
 
tidyr::pivot_longer(data, tidyselect::starts_with("var"), names_to = "year", names_prefix = "var", names_transform = list(year = as.integer), values_to = "var")
tidyr 		Wide to long
Differs from Stata: Converts numeric suffixes to integer year values, matching reshape long without string. pivot_longer() returns a text year without names_transform and does not enforce reshape i()/j() uniqueness constraints.
Stata reference · R reference 
reshape wide var, i(id) j(year)
 
pivot_wider(data, names_from = year, values_from = var, names_prefix = "var")
tidyr 		Long to wide
Differs from Stata: Duplicate keys can create list-columns in pivot_wider(); Stata reshape wide rejects duplicate i()/j() combinations. Validate keys first.
Stata reference · R reference 
reshape long var, i(id) j(year)
 
stats::reshape(data, direction = "long", varying = list(c("var2000", "var2001")), v.names = "var", timevar = "year", times = c(2000, 2001), idvar = "id")
stats 		Reshape repeated measurements to long form with base R
Differs from Stata: Explicitly list the wide columns and time values. This fixed workflow uses var2000 and var2001; adapt those names to your data.
Stata reference · R reference 
reshape wide var, i(id) j(year)
 
stats::reshape(data, direction = "wide", v.names = "var", timevar = "year", idvar = "id", sep = "")
stats 		Reshape repeated measurements to wide form with base R
Differs from Stata: Requires one row per id/year combination. R row ordering and attributes differ from Stata; compare keys and values explicitly.
Stata reference · R reference 
xpose, clear
 
tibble::as_tibble(t(data), .name_repair = "unique")
tibble 		Transpose the data
Stata reference · R reference

Construct data

Stata R Purpose, qualifications and sources 
expand n
 
uncount(data, n)
tidyr 		Repeat each row n times
Differs from Stata: uncount() drops n afterwards; add .remove = FALSE to keep it. uncount() drops the count column by default and zero counts remove rows; Stata expand keeps the count variable and retains observations when the count is below one.
Stata reference · R reference 
fillin x1 x2
 
complete(data, x1, x2)
tidyr 		Add rows for every combination of values
Differs from Stata: expand(data, x1, x2) returns only the combinations; expand_grid(x1 = 1:5, x2 = c("a", "b")) builds them from scratch. complete() adds combinations without the _fillin indicator created by Stata fillin; factor levels and missing-key combinations also need explicit decisions.
Stata reference · R reference 
input x y
 
tibble(x = c(1, 2), y = c(3, 4))
tibble 		Type in a small dataset
Stata reference · R reference 
set obs 100
 
data <- tibble(id = 1:100)
tibble 		Start a dataset with a given number of rows
Stata reference · R reference

Strings and dates

Stata R Purpose, qualifications and sources 
generate z = char(code)
 
data$z <- vapply(data$code,intToUtf8,character(1))
base 		Create characters from ASCII codes
Differs from Stata: Restricted to ASCII codes 1–127; extended-byte encodings are not Unicode code points.
Stata reference · R reference 
generate t = clock(s, "YMDhms")
 
mutate(data, t = lubridate::ymd_hms(s))
lubridate 		Text to a date-time
Differs from Stata: R uses Date/POSIXct or tsibble calendar classes, not Stata numeric date/time counts from 1960; convert epochs and time zones when exchanging underlying numeric values.
Stata reference · R reference 
generate double z = clock(s, "YMDhms")
 
data$z <- as.POSIXct(data$s,format='%Y-%m-%d %H:%M:%S',tz='UTC')
base 		Parse a date-time with an explicit UTC timezone
Differs from Stata: POSIXct is seconds since 1970 with a timezone; Stata clock is milliseconds since 1960 without a timezone.
Stata reference · R reference 
generate double z = cofd(d)
 
data$z <- as.numeric(data$d-as.Date('1960-01-01'))*86400000
base 		Convert R dates to Stata midnight millisecond codes
Differs from Stata: Returns ordinary Stata %tc milliseconds since 1960, not a POSIXct object.
Stata reference · R reference 
generate d = date(s, "YMD")
 
mutate(data, d = lubridate::ymd(s))
lubridate 		Text to a date
Differs from Stata: R uses Date/POSIXct or tsibble calendar classes, not Stata numeric date/time counts from 1960; convert epochs and time zones when exchanging underlying numeric values.
Stata reference · R reference 
generate z = date(s, "DMY")
 
data$z <- as.Date(data$s,format='%d/%m/%Y')
base 		Parse day-month-year date strings
Differs from Stata: R Date objects use a different numeric origin from Stata; use the object for calendar operations, not raw-code comparison.
Stata reference · R reference 
generate z = date(s, "MDY")
 
data$z <- as.Date(data$s,format='%m/%d/%Y')
base 		Parse month-day-year date strings
Differs from Stata: R Date objects use a different numeric origin from Stata; use the object for calendar operations, not raw-code comparison.
Stata reference · R reference 
generate z = date(s, "YMD")
 
data$z <- as.Date(data$s,format='%Y-%m-%d')
base 		Parse year-month-day date strings
Differs from Stata: R Date objects use a different numeric origin from Stata; use the object for calendar operations, not raw-code comparison.
Stata reference · R reference 
generate dd = day(d)
 
mutate(data, dd = lubridate::day(d))
lubridate 		Day of the month
Stata reference · R reference 
generate z = day(d)
 
data$z <- as.integer(format(data$d,'%d'))
base 		Extract day-of-month numbers
Differs from Stata: Accepts an R Date column and returns the demonstrated calendar component; raw Stata date codes require conversion first.
Stata reference · R reference 
generate z = daysinmonth(d)
 
data$z <- as.integer(lubridate::days_in_month(data$d))
lubridate 		Get the number of days in each calendar month
Differs from Stata: R Date objects use a different numeric origin from Stata; use the object for calendar operations, not raw-code comparison.
Stata reference · R reference 
display c(current_date)
 
Sys.Date()
base 		Today's date
Stata reference · R reference 
generate z = dofc(milliseconds)
 
data$z <- as.Date(as.POSIXct(data$milliseconds/1000,origin='1960-01-01',tz='UTC'),tz='UTC')
base 		Convert Stata millisecond datetimes to dates
Differs from Stata: For ordinary Stata %tc values without leap-second encoding; UTC supplies a deterministic timezone.
Stata reference · R reference 
generate d = dofm(m)
 
mutate(data, d = as.Date(m))
tsibble 		First day of a monthly period
Differs from Stata: Here m is a tsibble::yearmonth() value. R uses Date/POSIXct or tsibble calendar classes, not Stata numeric date/time counts from 1960; convert epochs and time zones when exchanging underlying numeric values.
Stata reference · R reference 
generate z = dofm(m)
 
data$z <- as.Date(sprintf('%04d-%02d-01',1960L+data$m%/%12L,data$m%%12L+1L))
base 		Convert Stata monthly codes to month-start dates
Differs from Stata: R Date objects use a different numeric origin from Stata; use the object for calendar operations, not raw-code comparison.
Stata reference · R reference 
generate z = dofq(q)
 
data$z <- as.Date(sprintf('%04d-%02d-01',1960L+data$q%/%4L,(data$q%%4L)*3L+1L))
base 		Convert Stata quarterly codes to quarter-start dates
Differs from Stata: R Date objects use a different numeric origin from Stata; use the object for calendar operations, not raw-code comparison.
Stata reference · R reference 
generate z = dofy(year)
 
data$z <- as.Date(sprintf('%04d-01-01',data$year))
base 		Construct the first day of each year
Differs from Stata: R Date objects use a different numeric origin from Stata; use the object for calendar operations, not raw-code comparison.
Stata reference · R reference 
generate w = dow(d)
 
mutate(data, w = lubridate::wday(d) - 1)
lubridate 		Day of the week
wday() counts Sunday as 1; Stata's dow() counts Sunday as 0.
Stata reference · R reference 
generate z = dow(d)
 
data$z <- as.integer(format(data$d,'%w'))
base 		Extract weekday with Sunday coded zero
Differs from Stata: Accepts an R Date column and returns the demonstrated calendar component; raw Stata date codes require conversion first.
Stata reference · R reference 
generate z = doy(d)
 
data$z <- as.integer(format(data$d,'%j'))
base 		Extract day-of-year numbers
Differs from Stata: Accepts an R Date column and returns the demonstrated calendar component; raw Stata date codes require conversion first.
Stata reference · R reference 
generate z = firstdayofmonth(d)
 
data$z <- as.Date(format(data$d,'%Y-%m-01'))
base 		Find the first date of each calendar month
Differs from Stata: R Date objects use a different numeric origin from Stata; use the object for calendar operations, not raw-code comparison.
Stata reference · R reference 
format d %td
 
format(data$d, "%d%b%Y")
base 		Display a date
Dates keep their class; format() only changes how they print.
Stata reference · R reference 
generate age = (d2 - d1) / 365.25
 
mutate(data, age = lubridate::time_length(lubridate::interval(d1, d2), "years"))
lubridate 		Age or duration in years
Differs from Stata: time_length(interval(...), "years") uses calendar-year lengths; dividing days by 365.25 is a fixed-length approximation and can give different ages.
Stata reference · R reference 
generate z = halfyear(d)
 
data$z <- (as.integer(format(data$d,'%m'))-1L)%/%6L+1L
base 		Extract half-year numbers
Differs from Stata: Accepts an R Date column and returns the demonstrated calendar component; raw Stata date codes require conversion first.
Stata reference · R reference 
generate z = hh(t)
 
data$z <- as.integer(format(data$t,'%H',tz='UTC'))
base 		Extract the hour from a UTC timestamp
Differs from Stata: Input is an R POSIXct timestamp; convert raw Stata %tc numbers before extracting components.
Stata reference · R reference 
generate s = lower(name)
 
mutate(data, s = str_to_lower(name))
stringr 		Lower case
Differs from Stata: R Unicode, regular-expression and case rules can differ from non-ustr Stata string functions; Stata empty strings are missing whereas R distinguishes empty strings and NA.
Stata reference · R reference 
generate d = mdy(m, d, y)
 
mutate(data, d = lubridate::make_date(y, m, d))
lubridate 		Date from year, month and day
Differs from Stata: R uses Date/POSIXct or tsibble calendar classes, not Stata numeric date/time counts from 1960; convert epochs and time zones when exchanging underlying numeric values.
Stata reference · R reference 
generate z = mdy(month, day, year)
 
data$z <- as.Date(sprintf('%04d-%02d-%02d',data$year,data$month,data$day))
base 		Construct a date from year, month and day columns
Differs from Stata: R Date objects use a different numeric origin from Stata; use the object for calendar operations, not raw-code comparison.
Stata reference · R reference 
generate z = mm(t)
 
data$z <- as.integer(format(data$t,'%M',tz='UTC'))
base 		Extract the minute from a UTC timestamp
Differs from Stata: Input is an R POSIXct timestamp; uppercase %M means minute and lowercase %m means month.
Stata reference · R reference 
generate m = mofd(d)
 
mutate(data, m = tsibble::yearmonth(d))
tsibble 		Monthly period from a date
Differs from Stata: tsibble::yearquarter() is the quarterly version, like qofd(). R uses Date/POSIXct or tsibble calendar classes, not Stata numeric date/time counts from 1960; convert epochs and time zones when exchanging underlying numeric values.
Stata reference · R reference 
generate z = mofd(d)
 
data$z <- (as.integer(format(data$d,'%Y'))-1960L)*12L+as.integer(format(data$d,'%m'))-1L
base 		Convert Date values to Stata monthly codes
Differs from Stata: Returns numeric months since January 1960, explicitly using Stata monthly units.
Stata reference · R reference 
generate m = month(d)
 
mutate(data, m = lubridate::month(d))
lubridate 		Month of a date
Stata reference · R reference 
generate z = month(d)
 
data$z <- as.integer(format(data$d,'%m'))
base 		Extract month numbers
Differs from Stata: Accepts an R Date column and returns the demonstrated calendar component; raw Stata date codes require conversion first.
Stata reference · R reference 
generate q = qofd(d)
 
mutate(data, q = tsibble::yearquarter(d))
tsibble 		Quarterly period from a date
Differs from Stata: R uses Date/POSIXct or tsibble calendar classes, not Stata numeric date/time counts from 1960; convert epochs and time zones when exchanging underlying numeric values.
Stata reference · R reference 
generate z = qofd(d)
 
data$z <- (as.integer(format(data$d,'%Y'))-1960L)*4L+(as.integer(format(data$d,'%m'))-1L)%/%3L
base 		Convert Date values to Stata quarterly codes
Differs from Stata: Returns numeric quarters since 1960q1, explicitly using Stata quarterly units.
Stata reference · R reference 
generate z = quarter(d)
 
data$z <- (as.integer(format(data$d,'%m'))-1L)%/%3L+1L
base 		Extract calendar quarter numbers
Differs from Stata: Accepts an R Date column and returns the demonstrated calendar component; raw Stata date codes require conversion first.
Stata reference · R reference 
generate x = real(s)
 
mutate(data, x = as.numeric(s))
base 		Text to a number
Differs from Stata: destring() from tidystata also strips value labels. as.numeric() warns for invalid text, unlike Stata real(); factors must be converted through their text levels before numeric conversion.
Stata reference · R reference 
generate z = real(s)
 
valid <- grepl('^[+-]?[0-9]+([.][0-9]+)?$',data$s); data$z <- rep(NA_real_,length(valid)); data$z[valid] <- as.numeric(data$s[valid])
base 		Parse a controlled decimal grammar with missing for invalid text
Differs from Stata: Deliberately limited grammar avoids accidental warnings; scientific notation and tagged missing require extension.
Stata reference · R reference 
generate has = regexm(name, "^A")
 
mutate(data, has = str_detect(name, "^A"))
stringr 		Match a regular expression
Differs from Stata: R Unicode, regular-expression and case rules can differ from non-ustr Stata string functions; Stata empty strings are missing whereas R distinguishes empty strings and NA.
Stata reference · R reference 
generate z = regexm(s, "^[0-9]+$")
 
data$z <- as.integer(grepl('^[0-9]+$',data$s))
base 		Flag strings containing only ASCII digits
Differs from Stata: This simple shared regex is portable; Stata and R regex engines differ for advanced patterns.
Stata reference · R reference 
generate s = regexr(name, "[0-9]+", "")
 
mutate(data, s = str_replace(name, "[0-9]+", ""))
stringr 		Replace the first regular-expression match
Differs from Stata: str_replace_all() replaces every match, like ustrregexra(). R Unicode, regular-expression and case rules can differ from non-ustr Stata string functions; Stata empty strings are missing whereas R distinguishes empty strings and NA.
Stata reference · R reference 
generate z = regexr(s, "[0-9]+", "#")
 
data$z <- sub('[0-9]+','#',data$s)
base 		Replace the first numeric run in text
Differs from Stata: First-match regex replacement; advanced regex features differ across engines.
Stata reference · R reference 
generate z = regexreplaceall(s, "[0-9]+", "#")
 
data$z <- gsub('[0-9]+','#',data$s)
base 		Replace all numeric runs in text
Differs from Stata: StataNow regexreplaceall workflow with a simple ASCII pattern; older Stata versions use other functions.
Stata reference · R reference 
generate d = regexs(1) if regexm(name, "([0-9]+)")
 
mutate(data, d = str_extract(name, "[0-9]+"))
stringr 		Extract the matching text
Differs from Stata: str_extract() returns NA for nonmatches; Stata generate ... if leaves empty strings in those rows. regexs() also depends on the preceding regexm() match.
Stata reference · R reference 
generate z = regexs(1) if regexm(s, "([A-Z]+)-([0-9]+)")
 
data$z <- stringr::str_match(data$s,'([A-Z]+)-([0-9]+)')[,2]
stringr 		Extract the first capturing group
R returns the group directly without a prior global regex-match state; unmatched rows use NA.
Stata reference · R reference 
split name, parse(" ")
 
separate_wider_delim(data, name, delim = " ", names_sep = "_", too_few = "align_start")
tidyr 		Split text into several variables
Stata reference · R reference 
split s, parse("-") destring
 
data <- tidyr::separate_wider_delim(data,s,delim='-',names=c('s1','s2'),cols_remove=FALSE); data$s1 <- as.integer(data$s1); data$s2 <- as.integer(data$s2)
tidyr 		Split two numeric fields and convert their types
Differs from Stata: Requires exactly two clean integer fields; resulting column names s1 and s2 are fixed.
Stata reference · R reference 
split s, parse("-") generate(part)
 
parts <- strsplit(data$s,'-',fixed=TRUE); data$part1 <- vapply(parts,`[`,character(1),1); data$part2 <- vapply(parts,`[`,character(1),2)
base 		Split two-component delimited strings
Differs from Stata: Fixed two-field example with exactly one delimiter per row; generated stub part is fixed.
Stata reference · R reference 
generate z = ss(t)
 
data$z <- as.numeric(format(data$t,'%OS',tz='UTC'))
base 		Extract seconds from a UTC timestamp
Differs from Stata: Input is an R POSIXct timestamp; fractional-second precision depends on the platform and display settings.
Stata reference · R reference 
generate s = string(x)
 
mutate(data, s = as.character(x))
base 		Number to text
Differs from Stata: tidystata's decode() uses value labels for labelled data; tostring() keeps the codes. as.character() follows R number formatting and NA handling; Stata string() uses numeric display formats and "." for numeric missing.
Stata reference · R reference 
generate s = stritrim(name)
 
mutate(data, s = str_squish(name))
stringr 		Collapse repeated spaces
Differs from Stata: str_squish() also removes leading and trailing spaces.
Stata reference · R reference 
generate z = stritrim(s)
 
data$z <- gsub(' +',' ',data$s)
base 		Collapse repeated internal ASCII spaces
Stata reference · R reference 
generate n = strlen(name)
 
mutate(data, n = str_length(name))
stringr 		Length of a string
Differs from Stata: R Unicode, regular-expression and case rules can differ from non-ustr Stata string functions; Stata empty strings are missing whereas R distinguishes empty strings and NA.
Stata reference · R reference 
generate z = strlen(s)
 
data$z <- nchar(data$s,type='bytes')
base 		Count bytes rather than Unicode characters
Differs from Stata: Byte length depends on encoding; this fixture is explicitly UTF-8.
Stata reference · R reference 
generate s = strlower(name)
 
mutate(data, s = str_to_lower(name))
stringr 		Lower case
Differs from Stata: R Unicode, regular-expression and case rules can differ from non-ustr Stata string functions; Stata empty strings are missing whereas R distinguishes empty strings and NA.
Stata reference · R reference 
generate z = strlower(s)
 
data$z <- tolower(data$s)
base 		Lowercase ASCII text
Stata reference · R reference 
generate z = strltrim(s)
 
data$z <- sub('^ +','',data$s)
base 		Trim leading ASCII spaces only
Stata reference · R reference 
generate m = strmatch(name, "A*")
 
mutate(data, m = grepl(glob2rx("A*"), name))
utils 		Match a wildcard pattern
Differs from Stata: R Unicode, regular-expression and case rules can differ from non-ustr Stata string functions; Stata empty strings are missing whereas R distinguishes empty strings and NA.
Stata reference · R reference 
generate z = strmatch(s, "a*")
 
data$z <- as.integer(grepl(utils::glob2rx('a*'),data$s))
base 		Match a simple wildcard pattern
Differs from Stata: Demonstrates * wildcard only; review escaping for more complicated patterns.
Stata reference · R reference 
generate z = strofreal(x, "%06.0f")
 
data$z <- sprintf('%06.0f',data$x)
base 		Format a number as zero-padded text
Stata reference · R reference 
generate has = strpos(name, "a") > 0
 
mutate(data, has = str_detect(name, fixed("a")))
stringr 		Whether a string contains some text
Differs from Stata: R Unicode, regular-expression and case rules can differ from non-ustr Stata string functions; Stata empty strings are missing whereas R distinguishes empty strings and NA.
Stata reference · R reference 
generate z = strpos(s, "cat")
 
position <- regexpr('cat',data$s,fixed=TRUE); data$z <- ifelse(position<0,0L,as.integer(position))
base 		Locate a literal ASCII substring with zero for absence
Stata reference · R reference 
generate s = strproper(name)
 
mutate(data, s = str_to_title(name))
stringr 		Title case
Differs from Stata: R Unicode, regular-expression and case rules can differ from non-ustr Stata string functions; Stata empty strings are missing whereas R distinguishes empty strings and NA.
Stata reference · R reference 
generate z = strproper(s)
 
data$z <- stringr::str_to_title(data$s,locale='en')
stringr 		Capitalize words in English place names
Differs from Stata: English locale example; punctuation and language-specific capitalization can differ.
Stata reference · R reference 
generate z = strreverse(s)
 
data$z <- vapply(strsplit(data$s,''),function(z) paste(rev(z),collapse=''),character(1))
base 		Reverse plain ASCII text
Stata reference · R reference 
generate z = strrtrim(s)
 
data$z <- sub(' +$','',data$s)
base 		Trim trailing ASCII spaces only
Stata reference · R reference 
generate n = strtoname(s)
 
mutate(data, n = make.names(s))
base 		Turn text into a valid name
Differs from Stata: janitor::clean_names(data) cleans every variable name at once. make.names() allows dots and follows R reserved-word rules; strtoname() follows Stata allowed-character and length limits.
Stata reference · R reference 
generate z = strtoname(s)
 
data$z <- substr(gsub('[^A-Za-z0-9_]','_',data$s),1,32); data$z <- ifelse(grepl('^[0-9]',data$z),paste0('_',data$z),data$z)
base 		Convert short ASCII text to Stata-style names
Differs from Stata: Demonstrated short ASCII names only; R syntactic names and Stata reserved names follow different rules.
Stata reference · R reference 
generate s = strtrim(name)
 
mutate(data, s = str_trim(name))
stringr 		Remove leading and trailing spaces
Differs from Stata: R Unicode, regular-expression and case rules can differ from non-ustr Stata string functions; Stata empty strings are missing whereas R distinguishes empty strings and NA.
Stata reference · R reference 
generate z = strtrim(s)
 
data$z <- trimws(data$s,whitespace=' ')
base 		Trim leading and trailing ASCII spaces
Stata reference · R reference 
generate s = strupper(name)
 
mutate(data, s = str_to_upper(name))
stringr 		Upper case
Differs from Stata: R Unicode, regular-expression and case rules can differ from non-ustr Stata string functions; Stata empty strings are missing whereas R distinguishes empty strings and NA.
Stata reference · R reference 
generate z = strupper(s)
 
data$z <- toupper(data$s)
base 		Uppercase ASCII text
Stata reference · R reference 
generate s = subinstr(name, "a", "b", .)
 
mutate(data, s = str_replace_all(name, fixed("a"), "b"))
stringr 		Replace text
Differs from Stata: stringr treats patterns as regular expressions; fixed() matches the text literally, as subinstr() does. R Unicode, regular-expression and case rules can differ from non-ustr Stata string functions; Stata empty strings are missing whereas R distinguishes empty strings and NA.
Stata reference · R reference 
generate z = subinstr(s, "cat", "dog", .)
 
data$z <- gsub('cat','dog',data$s,fixed=TRUE)
base 		Replace every literal occurrence
Stata reference · R reference 
generate z = subinstr(s, "cat", "dog", 1)
 
data$z <- sub('cat','dog',data$s,fixed=TRUE)
base 		Replace only the first literal occurrence
Stata reference · R reference 
generate z = subinword(s, "cat", "dog", .)
 
data$z <- gsub('\\bcat\\b','dog',data$s,perl=TRUE)
base 		Replace a complete word in simple ASCII text
Differs from Stata: Uses regular-expression word boundaries; delimiters and Unicode rules can differ from Stata.
Stata reference · R reference 
generate s = substr(name, 1, 3)
 
mutate(data, s = str_sub(name, 1, 3))
stringr 		Part of a string
Differs from Stata: R Unicode, regular-expression and case rules can differ from non-ustr Stata string functions; Stata empty strings are missing whereas R distinguishes empty strings and NA.
Stata reference · R reference 
generate z = substr(s, -3, .)
 
data$z <- stringr::str_sub(data$s,-3)
stringr 		Extract the last three ASCII characters
Stata reference · R reference 
generate z = substr(s, 1, 3)
 
data$z <- substr(data$s,1,3)
base 		Extract an ASCII prefix by length
Stata reference · R reference 
generate z = substr(s, 3, 2)
 
data$z <- substr(data$s,3,4)
base 		Extract ASCII text using start and length
Stata takes a length; base R substr takes an inclusive ending position.
Stata reference · R reference 
generate z = uchar(code)
 
data$z <- vapply(data$code,intToUtf8,character(1))
base 		Create characters from Unicode code points
Stata reference · R reference 
generate s = upper(name)
 
mutate(data, s = str_to_upper(name))
stringr 		Upper case
Differs from Stata: R Unicode, regular-expression and case rules can differ from non-ustr Stata string functions; Stata empty strings are missing whereas R distinguishes empty strings and NA.
Stata reference · R reference 
generate z = ustrlen(s)
 
data$z <- nchar(data$s,type='chars')
base 		Count Unicode code points
Counts code points, not visible grapheme clusters.
Stata reference · R reference 
generate z = ustrlower(s)
 
data$z <- stringr::str_to_lower(data$s,locale='en')
stringr 		Lowercase Unicode text with an explicit locale
Differs from Stata: The locale is explicitly English; choose the relevant locale for language-sensitive casing.
Stata reference · R reference 
generate z = ustrnormalize(s, "nfc")
 
data$z <- stringi::stri_trans_nfc(data$s)
stringi 		Normalize canonically equivalent Unicode text
Stata reference · R reference 
generate z = ustrnormalize(s, "nfkc")
 
data$z <- stringi::stri_trans_nfkc(data$s)
stringi 		Normalize compatibility Unicode characters
Stata reference · R reference 
generate z = ustrpos(s, "猫")
 
position <- regexpr('\u732b',data$s,fixed=TRUE); data$z <- ifelse(position<0,0L,as.integer(position))
base 		Locate a Unicode substring by code-point position
Stata reference · R reference 
generate has = ustrregexm(name, "^a", 1)
 
mutate(data, has = str_detect(name, regex("^a", ignore_case = TRUE)))
stringr 		Unicode-aware regular expression, ignoring case
Differs from Stata: stringr is Unicode-aware throughout, so the u- forms of Stata's string functions need no special version. R Unicode, regular-expression and case rules can differ from non-ustr Stata string functions; Stata empty strings are missing whereas R distinguishes empty strings and NA.
Stata reference · R reference 
generate z = ustrregexm(s, "\p{L}+")
 
data$z <- as.integer(stringr::str_detect(data$s,'\\p{L}+'))
stringr 		Detect Unicode letters
Differs from Stata: ICU regex workflow; missing strings in R remain NA rather than empty Stata strings.
Stata reference · R reference 
generate z = ustrregexra(s, "[0-9]+", "#")
 
data$z <- stringr::str_replace_all(data$s,'[0-9]+','#')
stringr 		Replace every ICU regex match
Stata reference · R reference 
generate z = ustrregexrf(s, "[0-9]+", "#")
 
data$z <- stringr::str_replace(data$s,'[0-9]+','#')
stringr 		Replace the first ICU regex match
Stata reference · R reference 
generate z = ustrreverse(s)
 
data$z <- stringi::stri_reverse(data$s)
stringi 		Reverse Unicode code-point order
Differs from Stata: Combining marks and grapheme clusters may require a different user-facing reversal rule.
Stata reference · R reference 
generate z = ustrtrim(s)
 
data$z <- stringr::str_trim(data$s)
stringr 		Trim Unicode whitespace
Stata reference · R reference 
generate z = ustrupper(s)
 
data$z <- stringr::str_to_upper(data$s,locale='en')
stringr 		Uppercase Unicode text with an explicit locale
Differs from Stata: Explicit English locale; locale-specific casing can differ from Stata defaults.
Stata reference · R reference 
generate z = usubstr(s, 2, 1)
 
data$z <- substr(data$s,2,2)
base 		Extract one Unicode code point by position
R character indexing follows code points; this does not count grapheme clusters.
Stata reference · R reference 
generate w = word(name, 1)
 
mutate(data, w = stringr::word(name, 1))
stringr 		The first word of a string
Differs from Stata: R Unicode, regular-expression and case rules can differ from non-ustr Stata string functions; Stata empty strings are missing whereas R distinguishes empty strings and NA.
Stata reference · R reference 
generate z = word(s, 2)
 
data$z <- vapply(strsplit(trimws(data$s),' +'),function(z) z[2],character(1))
base 		Extract the second simple whitespace word
Differs from Stata: Simple unquoted words with at least two tokens; Stata quoted-word parsing is not reproduced.
Stata reference · R reference 
generate n = wordcount(name)
 
mutate(data, n = str_count(name, "\\S+"))
stringr 		Number of words
Differs from Stata: R Unicode, regular-expression and case rules can differ from non-ustr Stata string functions; Stata empty strings are missing whereas R distinguishes empty strings and NA.
Stata reference · R reference 
generate z = wordcount(s)
 
data$z <- ifelse(trimws(data$s)=='',0L,lengths(strsplit(trimws(data$s),' +')))
base 		Count simple whitespace-separated words
Differs from Stata: Unquoted ASCII words only; quoted multiword phrases need a parser.
Stata reference · R reference 
generate y = year(d)
 
mutate(data, y = lubridate::year(d))
lubridate 		Year of a date
lubridate::month(), day() and wday() extract the other parts.
Stata reference · R reference 
generate z = year(d)
 
data$z <- as.integer(format(data$d,'%Y'))
base 		Extract calendar years
Differs from Stata: Accepts an R Date column and returns the demonstrated calendar component; raw Stata date codes require conversion first.
Stata reference · R reference 
generate z = ym(year, month)
 
data$z <- (data$year-1960L)*12L+data$month-1L
base 		Construct Stata monthly codes from year and month
Differs from Stata: Requires valid integer months 1–12; output uses Stata months since January 1960.
Stata reference · R reference 
generate z = yq(year, quarter)
 
data$z <- (data$year-1960L)*4L+data$quarter-1L
base 		Construct Stata quarterly codes from year and quarter
Differs from Stata: Requires valid integer quarters 1–4; output uses quarters since 1960q1.
Stata reference · R reference

Loops and programming

Stata R Purpose, qualifications and sources 
ado update
 
update.packages()
utils 		Update installed packages
Stata reference · R reference 
capture command
 
capture(expression)
tidystata 		Run code and keep its result without stopping on an error
Differs from Stata: Returns the expression's result on success; on error, emits a failure message and returns invisible NULL. quiet = TRUE suppresses the notice, warnings and messages, but keeps printed output. Use capture(data <- update(data)) to preserve an existing object if the update fails; data <- capture(update(data)) assigns NULL on error. Separate assignments in a captured block keep the last completed update; a single assignment around a failed pipeline keeps the original target. Earlier effects are not rolled back. This adapts Stata capture, which suppresses all output.
Stata reference · R reference 
clear all
 
rm(list = ls())
base 		Remove every object
Restarting R (Session > Restart R in RStudio) gives a truly clean start.
Stata reference · R reference 
cls
 
cat("\014")
base 		Clear the console
In RStudio, Ctrl+L does the same.
Stata reference · R reference 
display _b[x]
 
coef(fit)["x"]
stats 		A stored coefficient
Stata reference · R reference 
display _se[x]
 
sqrt(diag(vcov(fit)))["x"]
stats 		A stored standard error
Stata reference · R reference 
ereturn list
 
broom::glance(fit)
broom 		Model statistics
broom::tidy(fit) gives the coefficient table, and names(fit) lists every stored part.
Stata reference · R reference 
foreach v of varlist v1-v5 { generate `v'_sq = `v'^2 }
 
mutate(data, across(v1:v5, ~ .x^2, .names = "{col}_sq"))
dplyr 		Same calculation for several variables
across() replaces most variable loops; for other loops R has for (v in names) { ... }.
Stata reference · R reference 
forvalues i = 1/5 { generate v`i'_sq = v`i'^2 if v`i' > 0 }
 
mutate(data, across(v1:v5, ~ if_else(.x > 0, .x^2, NA), .names = "{col}_sq"))
dplyr 		Loop with a condition
Stata reference · R reference 
forvalues i = 6/10 { forvalues j = 3/5 { replace x_`i'_`j' = x_`i'_`j' + 1 } }
 
mutate(data, across(matches("^x_([6-9]|10)_[3-5]$"), ~ .x + 1))
dplyr 		Loop over variables numbered in their names
A regular expression picks the names. varlist("x_[6-10]_*") works when only one number range is needed.
Stata reference · R reference 
frame create other
 
other <- tibble()
tibble 		Hold another dataset
R keeps any number of data frames as separate objects, so frames are built in.
Stata reference · R reference 
if `x' > 0 { display "positive" }
 
if (x > 0) print("positive")
base 		Run code only if a condition holds
Quote it: findit("if"), because if is a reserved word in R.
Stata reference · R reference 
local vars x1 x2 x3
 
vars <- c("x1", "x2", "x3")
base 		Store a list of variable names
Use it with all_of(): select(data, all_of(vars)) or across(all_of(vars), ...).
Stata reference · R reference 
macro list
 
ls()
base 		List defined objects
Stata reference · R reference 
matrix A = (1, 2 \ 3, 4)
 
A <- matrix(c(1, 2, 3, 4), nrow = 2, byrow = TRUE)
base 		Create a matrix
Stata reference · R reference 
preserve
 
backup <- data
base 		Keep a copy to return to
R never changes data in place, so keep the original object and assign results to a new name; restore is data <- backup.
Stata reference · R reference 
program define myprog
 
myprog <- function(data) { summ(data) }
base 		Write your own command
Stata reference · R reference 
quietly regress y x
 
fit <- lm(y ~ x, data = data)
stats 		Run without printing
Assigning a result never prints it; print(fit) or summary(fit) shows it later.
Stata reference · R reference 
return list
 
str(summ(data, x))
tidystata 		See the results a command stored
R results are ordinary objects: assign them and pick out parts, as in s <- summ(data, x); s$mean.
Stata reference · R reference 
scalar a = 3
 
a <- 3
base 		Store a single number
Stata reference · R reference 
set seed 2024
 
set.seed(2024)
base 		Make random numbers reproducible
Differs from Stata: set_seed(.fn, seed) sets the seed inside a pipe; see the sample rows. R and Stata use different random-number streams; the same numeric seed does not reproduce the same draws or selected rows.
Stata reference · R reference 
set trace on
 
traceback()
base 		Find where an error happened
debug(f) or browser() steps through a function line by line.
Stata reference · R reference 
tempfile f
 
f <- tempfile(fileext = ".rds")
base 		A temporary file
Temporary variables are not needed: intermediate columns can be dropped, or kept as separate objects.
Stata reference · R reference 
timer on 1
 
timer(expression)
tidystata 		Time how long code takes
Stata reference · R reference 
version 18
 
R.version.string
base 		Version control
renv::snapshot() records package versions so a project can be reproduced later.
Stata reference · R reference 
which summarize
 
getAnywhere("summ")
utils 		Where a command comes from
Stata reference · R reference 
while `i' < 10 { local i = `i' + 1 }
 
while (i < 10) { i <- i + 1 }
base 		Repeat while a condition holds
Stata reference · R reference

Models and tables

Stata R Purpose, qualifications and sources 
areg y x, absorb(id)
 
fixest::feols(y ~ x | id, data = data)
fixest 		Regression absorbing one categorical variable
Differs from Stata: Match fixest covariance small-sample corrections and singleton handling with Stata; the same formula does not guarantee identical standard errors or retained observations.
Stata reference · R reference 
avplots
 
car::avPlots(fit)
car 		Added-variable plots
Stata reference · R reference 
bootstrap, reps(500): regress y x
 
lmtest::coeftest(fit, vcov = sandwich::vcovBS(fit, R = 500))
sandwich 		Bootstrap standard errors
Differs from Stata: The R code estimates covariance for an existing fit; resampling units, centering and t/normal inference must match the Stata bootstrap/jackknife prefix options.
Stata reference · R reference 
cluster kmeans x1 x2, k(3)
 
kmeans(na.omit(select(data, x1, x2)), centers = 3)
stats 		K-means clustering
Differs from Stata: R defaults to Hartigan-Wong k-means and random initial centers; Stata initialization and iteration defaults differ. Control the starting centers to compare clusters.
Stata reference · R reference 
csdid y, ivar(id) time(year) gvar(first_treat)
 
did::att_gt(yname = "y", tname = "year", idname = "id", gname = "first_treat", data = data)
did 		Staggered difference in differences
Differs from Stata: Match control-group, anticipation and estimation-method settings; csdid and att_gt() defaults and aggregation/inference need not coincide.
Stata reference · R reference 
didregress (y) (treated), group(id) time(year)
 
fixest::feols(y ~ treated | id + year, data = data)
fixest 		Difference in differences (Stata 17+)
Differs from Stata: The TWFE coefficient alone omits didregress default group-clustered standard errors and diagnostics; specify clustering and verify the DID design.
Stata reference · R reference 
estat hettest
 
lmtest::bptest(fit, ~ fitted(fit), studentize = FALSE)
lmtest 		Breusch-Pagan test for heteroskedasticity
These options match Stata's default test against the fitted values.
Stata reference · R reference 
estat ic
 
c(AIC = AIC(fit), BIC = BIC(fit))
stats 		Information criteria
Differs from Stata: Compare the same fitted sample and likelihood constants; R classes and Stata can count scale parameters and likelihood terms differently.
Stata reference · R reference 
estat ovtest
 
lmtest::resettest(fit, power = 2:4)
lmtest 		RESET test for omitted variables
Stata reference · R reference 
estat vce
 
vcov(fit)
stats 		Variance-covariance matrix
Stata reference · R reference 
estimates store m1
 
m1 <- lm(y ~ x, data = data)
stats 		Keep a model for later
A fitted model is an ordinary object; keep as many as you like under different names.
Stata reference · R reference 
estimates table m1 m2
 
modelsummary::modelsummary(list(m1, m2), output = "markdown")
modelsummary 		Compare models side by side
Stata reference · R reference 
eststo m1: regress y x
 
m1 <- lm(y ~ x, data = data)
stats 		Store a model for a table
Stata reference · R reference 
esttab m1 m2
 
modelsummary::modelsummary(list(m1, m2))
modelsummary 		Regression table
Stata reference · R reference 
etable
 
modelsummary::modelsummary(fit)
modelsummary 		Table of the last model (Stata 17+)
Stata reference · R reference 
glm y x, family(binomial)
 
glm(y ~ x, family = binomial, data = data)
stats 		Generalised linear model
Stata reference · R reference 
heckman y x, select(s = z)
 
sampleSelection::selection(s ~ z, y ~ x, data = data)
sampleSelection 		Heckman selection model
Stata reference · R reference 
ivreg2 y x1 (x2 = z), robust
 
fixest::feols(y ~ x1 | x2 ~ z, data = data, vcov = "hetero")
fixest 		Instrumental variables with robust errors
Differs from Stata: fixest::fitstat(fit, ~ ivf + ivwald) reports first-stage statistics. Match fixest covariance small-sample corrections and singleton handling with Stata; the same formula does not guarantee identical standard errors or retained observations.
Stata reference · R reference 
ivregress 2sls y x1 (x2 = z)
 
fixest::feols(y ~ x1 | x2 ~ z, data = data)
fixest 		Instrumental variables (2SLS)
Differs from Stata: Match fixest covariance small-sample corrections and singleton handling with Stata; the same formula does not guarantee identical standard errors or retained observations.
Stata reference · R reference 
jackknife: regress y x
 
lmtest::coeftest(fit, vcov = sandwich::vcovJK(fit))
sandwich 		Jackknife standard errors
Differs from Stata: The R code estimates covariance for an existing fit; resampling units, centering and t/normal inference must match the Stata bootstrap/jackknife prefix options.
Stata reference · R reference 
lincom x1 - x2
 
marginaleffects::hypotheses(fit, "x1 - x2 = 0")
marginaleffects 		Linear combination of coefficients
Differs from Stata: hypotheses() uses the fitted R covariance and degrees-of-freedom defaults; Stata can use a different t, F or normal reference distribution.
Stata reference · R reference 
logistic y x
 
broom::tidy(glm(y ~ x, family = binomial, data = data), exponentiate = TRUE, conf.int = TRUE)
broom 		Logit model reported as odds ratios
Differs from Stata: broom confidence intervals are model-specific; glm profile-likelihood intervals can differ from Stata default Wald intervals, even when odds ratios agree.
Stata reference · R reference 
logit y x1 x2
 
glm(y ~ x1 + x2, family = binomial, data = data)
stats 		Logit model
Stata reference · R reference 
logit y x, or
 
broom::tidy(glm(y ~ x, family = binomial, data = data), exponentiate = TRUE)
broom 		Odds ratios
Differs from Stata: broom confidence intervals are model-specific; glm profile-likelihood intervals can differ from Stata default Wald intervals, even when odds ratios agree.
Stata reference · R reference 
lrtest m1 m2
 
lmtest::lrtest(m1, m2)
lmtest 		Likelihood-ratio test
Differs from Stata: Compare the same fitted sample and likelihood constants; R classes and Stata can count scale parameters and likelihood terms differently.
Stata reference · R reference 
margins group
 
marginaleffects::avg_predictions(fit, by = "group")
marginaleffects 		Predictive margins by group
Differs from Stata: by = "group" averages within observed groups; margins group sets each level for every observation before averaging. Use counterfactual predictions for that Stata estimand.
Stata reference · R reference 
margins, dydx(*)
 
marginaleffects::avg_slopes(fit)
marginaleffects 		Average marginal effects
Differs from Stata: avg_slopes() averages model derivatives; discrete factor changes, weights and covariance must match margins and the fitted-model context.
Stata reference · R reference 
marginsplot
 
marginaleffects::plot_predictions(fit, condition = "x")
marginaleffects 		Plot predictive margins
Stata reference · R reference 
melogit y x || id:
 
lme4::glmer(y ~ x + (1 | id), data = data, family = binomial)
lme4 		Multilevel logit
Differs from Stata: glmer() defaults to Laplace integration (nAGQ = 1); melogit defaults to adaptive quadrature, which can change estimates and likelihoods.
Stata reference · R reference 
mi estimate: regress y x
 
mice::pool(with(imp, lm(y ~ x)))
mice 		Fit and combine models across imputations
Differs from Stata: Rubin pooling still requires matching complete-data degrees of freedom and small-sample adjustments to mi estimate.
Stata reference · R reference 
mi impute chained (regress) x1 x2 = y
 
imp <- mice::mice(data, m = 20)
mice 		Multiple imputation
Differs from Stata: mice() chooses methods by R column type, often predictive mean matching; it does not default to the specified chained linear regressions. Match methods, predictors and m explicitly.
Stata reference · R reference 
mixed y x || id:
 
lme4::lmer(y ~ x + (1 | id), data = data)
lme4 		Multilevel (mixed) linear model
Differs from Stata: lme4 reports no p-values; lmerTest::lmer() adds them. lmer() defaults to REML; Stata mixed defaults to ML. Use REML = FALSE for ML and select a separate method for fixed-effect p-values.
Stata reference · R reference 
mlogit y x1 x2
 
nnet::multinom(y ~ x1 + x2, data = data)
nnet 		Multinomial logit
Differs from Stata: multinom() needs a factor y for multinomial classification; reference outcome categories and optimizer defaults can differ from Stata mlogit.
Stata reference · R reference 
nbreg y x
 
MASS::glm.nb(y ~ x, data = data)
MASS 		Negative binomial model
Differs from Stata: fixest::fenegbin() adds fixed effects. glm.nb() uses NB2 like default nbreg, but theta is the reciprocal of Stata alpha and likelihood/optimization reporting differs.
Stata reference · R reference 
nlcom _b[x1] / _b[x2]
 
marginaleffects::hypotheses(fit, "x1 / x2 = 0")
marginaleffects 		Nonlinear combination of coefficients
Differs from Stata: Standard errors use the delta method, as in Stata. hypotheses() uses the fitted R covariance and degrees-of-freedom defaults; Stata can use a different t, F or normal reference distribution.
Stata reference · R reference 
oaxaca y x1 x2, by(group)
 
oaxaca::oaxaca(y ~ x1 + x2 | group, data = data)
oaxaca 		Blinder-Oaxaca decomposition
Differs from Stata: oaxaca() reports several reference-weight decompositions; explicitly select the reference coefficients and variance method used by Stata oaxaca.
Stata reference · R reference 
ologit y x1 x2
 
MASS::polr(factor(y) ~ x1 + x2, data = data, Hess = TRUE)
MASS 		Ordered logit
Differs from Stata: Hess = TRUE lets summary() show standard errors. polr() uses ordered factor levels and a threshold parameterization; set the order explicitly and check the cutpoint sign convention against Stata.
Stata reference · R reference 
oprobit y x1 x2
 
MASS::polr(factor(y) ~ x1 + x2, data = data, method = "probit", Hess = TRUE)
MASS 		Ordered probit
Differs from Stata: polr() uses ordered factor levels and a threshold parameterization; set the order explicitly and check the cutpoint sign convention against Stata.
Stata reference · R reference 
outreg2 using table.doc
 
modelsummary::modelsummary(list(m1, m2), output = "table.docx")
modelsummary 		Regression table in a Word file
Stata reference · R reference 
poisson y x1 x2
 
glm(y ~ x1 + x2, family = poisson, data = data)
stats 		Poisson model
Stata reference · R reference 
ppmlhdfe y x, absorb(id year)
 
fixest::fepois(y ~ x | id + year, data = data)
fixest 		Poisson model with fixed effects
Differs from Stata: Poisson fixed-effect coefficient targets agree, but separation/singleton removal and covariance corrections can differ; check the retained sample and uncertainty.
Stata reference · R reference 
predict yhat
 
mutate(data, yhat = predict(fit, newdata = data))
stats 		Fitted values
Differs from Stata: residuals(fit) gives residuals for the rows used in the model. predict() depends on the R model: response/link scale, residual definition and prediction sample must match the Stata postestimation context.
Stata reference · R reference 
predict p, pr
 
mutate(data, p = predict(fit, newdata = data, type = "response"))
stats 		Predicted probabilities
Differs from Stata: predict() depends on the R model: response/link scale, residual definition and prediction sample must match the Stata postestimation context.
Stata reference · R reference 
predict r, residuals
 
mutate(data, r = y - predict(fit, newdata = data))
stats 		Residuals for every row
Differs from Stata: predict() depends on the R model: response/link scale, residual definition and prediction sample must match the Stata postestimation context.
Stata reference · R reference 
probit y x1 x2
 
glm(y ~ x1 + x2, family = binomial(link = "probit"), data = data)
stats 		Probit model
Stata reference · R reference 
putdocx begin
 
flextable::save_as_docx(flextable::flextable(results), path = "results.docx")
flextable 		Write a table to Word
officer builds whole Word documents.
Stata reference · R reference 
putexcel set results.xlsx
 
writexl::write_xlsx(results, "results.xlsx")
writexl 		Write results to Excel
openxlsx2 gives cell-level control over formatting.
Stata reference · R reference 
qreg y x
 
quantreg::rq(y ~ x, tau = 0.5, data = data)
quantreg 		Median (quantile) regression
Differs from Stata: rq() default standard-error methods depend on sample size and can differ from Stata qreg; choose the covariance/inference method explicitly.
Stata reference · R reference 
rdrobust y x, c(0)
 
rdrobust::rdrobust(data$y, data$x, c = 0)
rdrobust 		Regression discontinuity
Differs from Stata: The R and Stata implementations share the estimator; version, bandwidth, mass-point and variance settings must also match for finite-sample output.
Stata reference · R reference 
reghdfe y x, absorb(id year)
 
fixest::feols(y ~ x | id + year, data = data)
fixest 		High-dimensional fixed effects
Differs from Stata: Match fixest covariance small-sample corrections and singleton handling with Stata; the same formula does not guarantee identical standard errors or retained observations.
Stata reference · R reference 
regress y x1 x2
 
fit <- lm(y ~ x1 + x2, data = data)
stats 		Linear regression
summary(fit) shows the coefficient table; broom::tidy(fit) returns it as a data frame.
Stata reference · R reference 
regress y i.group c.x1##c.x2
 
lm(y ~ factor(group) + x1 * x2, data = data)
stats 		Factor variables and interactions
Differs from Stata: factor() plays the role of i.; x1 * x2 is c.x1##c.x2 and x1:x2 is c.x1#c.x2. factor() chooses its reference from R level order; Stata i.group uses its numeric base level. Set factor levels and contrasts explicitly.
Stata reference · R reference 
regress y ib3.group
 
lm(y ~ relevel(factor(group), ref = "3"), data = data)
stats 		Choose the base category
Stata reference · R reference 
regress y x [pweight = wt]
 
fixest::feols(y ~ x, data = data, weights = ~wt, vcov = "hetero")
fixest 		Regression with sampling weights
Differs from Stata: For full survey designs with strata and clusters, see svyset. Match fixest covariance small-sample corrections and singleton handling with Stata; the same formula does not guarantee identical standard errors or retained observations.
Stata reference · R reference 
regress y x1 x2, robust
 
fixest::feols(y ~ x1 + x2, data = data, vcov = "hetero")
fixest 		Heteroskedasticity-robust standard errors
Differs from Stata: Match fixest covariance small-sample corrections and singleton handling with Stata; the same formula does not guarantee identical standard errors or retained observations.
Stata reference · R reference 
regress y x1 x2, vce(cluster id)
 
fixest::feols(y ~ x1 + x2, data = data, cluster = ~id)
fixest 		Clustered standard errors
Differs from Stata: Match fixest covariance small-sample corrections and singleton handling with Stata; the same formula does not guarantee identical standard errors or retained observations.
Stata reference · R reference 
rreg y x
 
MASS::rlm(y ~ x, data = data)
MASS 		Robust regression
Differs from Stata: rlm() uses Huber M-estimation; Stata's rreg combines Huber and biweight steps, so estimates can differ slightly. rlm() defaults to Huber M-estimation; rreg screens outliers and combines Huber and biweight iterations, so estimates and inference can differ.
Stata reference · R reference 
rvfplot
 
plot(fit, which = 1)
stats 		Residuals against fitted values
Stata reference · R reference 
sem (y <- x1 x2)
 
lavaan::sem("y ~ x1 + x2", data = data)
lavaan 		Structural equation model
Differs from Stata: lavaan treats observed exogenous predictors as fixed by default; match fixed.x and missing-data likelihood settings with Stata sem.
Stata reference · R reference 
synth y x1 x2, trunit(1) trperiod(2000)
 
Synth::synth(data.prep.obj = prep)
Synth 		Synthetic control
Synth::dataprep() builds prep from the panel first; tidysynth offers a tidier interface.
Stata reference · R reference 
teffects ipw (y) (t x1 x2)
 
WeightIt::weightit(t ~ x1 + x2, data = data, method = "glm")
WeightIt 		Inverse-probability weighting
Stata reference · R reference 
teffects psmatch (y) (t x1 x2)
 
MatchIt::matchit(t ~ x1 + x2, data = data, method = "nearest")
MatchIt 		Propensity-score matching
Estimate the effect on the matched data, for example lm(y ~ t, data = MatchIt::match.data(m), weights = weights).
Stata reference · R reference 
test x1 = x2
 
marginaleffects::hypotheses(fit, "x1 = x2")
marginaleffects 		Test a hypothesis about coefficients
Differs from Stata: hypotheses() uses the fitted R covariance and degrees-of-freedom defaults; Stata can use a different t, F or normal reference distribution.
Stata reference · R reference 
tobit y x, ll(0)
 
AER::tobit(y ~ x, left = 0, data = data)
AER 		Censored (tobit) regression
Stata reference · R reference 
vif
 
car::vif(fit)
car 		Variance inflation factors
Differs from Stata: car::vif() gives generalized VIFs for multi-degree-of-freedom terms; Stata estat vif reports regressor-column VIFs, so factor-term tables differ.
Stata reference · R reference 
xtabond y x
 
plm::pgmm(y ~ lag(y, 1) + x | lag(y, 2:99), data = data, index = c("id", "year"))
plm 		Dynamic panel (Arellano-Bond)
Differs from Stata: pgmm() needs matching transformation, lag instruments, collapse and one/two-step choices; the formula alone does not reproduce xtabond default instruments or standard errors.
Stata reference · R reference 
xtlogit y x, fe
 
survival::clogit(y ~ x + strata(id), data = data)
survival 		Conditional (fixed-effects) logit
Differs from Stata: fixest::feglm(y ~ x | id, family = binomial) is faster but estimates unconditional fixed effects. clogit() targets conditional fixed-effects logit; exact likelihood, weights and covariance settings must match xtlogit, fe rather than unconditional dummy-variable logit.
Stata reference · R reference 
xtpoisson y x, fe
 
fixest::fepois(y ~ x | id, data = data)
fixest 		Fixed-effects Poisson
Differs from Stata: Poisson fixed-effect coefficient targets agree, but separation/singleton removal and covariance corrections can differ; check the retained sample and uncertainty.
Stata reference · R reference 
xtreg y x, fe
 
fixest::feols(y ~ x | id, data = data)
fixest 		Panel fixed effects
Differs from Stata: Match fixest covariance small-sample corrections and singleton handling with Stata; the same formula does not guarantee identical standard errors or retained observations.
Stata reference · R reference 
xtreg y x, re
 
plm::plm(y ~ x, data = data, index = c("id", "year"), model = "random")
plm 		Panel random effects
Differs from Stata: plm and xtreg, re default to different variance-component estimators; match random.method and the panel sample before comparing results.
Stata reference · R reference

Panel and time series

Stata R Purpose, qualifications and sources 
arima y, ar(1/2)
 
arima(data$y, order = c(2, 0, 0))
stats 		ARIMA model
Differs from Stata: R and Stata ARIMA initialization and intercept parameterizations differ; match the likelihood method and transform intercepts before comparing coefficients.
Stata reference · R reference 
carryforward x, replace
 
tidyr::fill(data, x, .direction = "down")
tidyr 		Carry the last observed value forward
Differs from Stata: group_by(data, id) first to fill within each panel unit. fill() crosses panel boundaries unless grouped first; group_by(id) is needed for within-unit carryforward.
Stata reference · R reference 
generate d_x = D.x
 
mutate(data, d_x = stata_diff(x, time = year), .by = id)
tidystata 		First difference within each panel
stata_diff(x, 2, time = year) is D2.x, the difference of differences.
Stata reference · R reference 
generate lead_x = F.x
 
mutate(data, lead_x = stata_lead(x, time = year), .by = id)
tidystata 		Lead within each panel
Gaps in year give missing values, as in Stata.
Stata reference · R reference 
generate lag_x = L.x
 
mutate(data, lag_x = stata_lag(x, time = year), .by = id)
tidystata 		Lag within each panel
Like Stata, gaps in year give missing values; dplyr::lag() shifts rows instead. Use stata_lag(x, 2, time = year) for L2.x and unit = "month" for monthly dates.
Stata reference · R reference 
generate lag2_x = L2.x
 
mutate(data, lag2_x = lag(x, 2, order_by = year), .by = id)
dplyr 		Two-period lag within each panel
Differs from Stata: lag(x, 2, order_by = year) shifts two observed rows; L2.x shifts two time units and returns missing across time gaps. Use stata_lag(x, 2, time = year) for that rule.
Stata reference · R reference 
tsfill
 
tsibble::fill_gaps(ts_data)
tsibble 		Add rows for missing time points
Differs from Stata: fill_gaps() fills each key's observed range by default; choose .full explicitly for common panel bounds and ensure the tsibble interval matches the Stata delta.
Stata reference · R reference 
tsline y
 
ggplot(data, aes(year, y)) + geom_line()
ggplot2 		Time-series line plot
Stata reference · R reference 
tsset year
 
tsibble::as_tsibble(data, index = year)
tsibble 		Declare time-series data
Stata reference · R reference 
xtdescribe
 
count(count(data, id, name = "periods"), periods)
dplyr 		How many periods each panel unit has
Stata reference · R reference 
xtline y
 
ggplot(data, aes(year, y)) + geom_line() + facet_wrap(~id)
ggplot2 		One line plot per panel unit
aes(group = id) without facets draws all units in one plot, like the overlay option.
Stata reference · R reference 
xtset id year
 
tsibble::as_tsibble(data, key = id, index = year)
tsibble 		Declare panel data
tidystata needs no declaration: stata_lag(x, time = year) with .by = id, and fixest takes panel.id = ~id + year.
Stata reference · R reference 
xtsum x
 
data |> mutate(m = mean(x, na.rm = TRUE), .by = id) |> summarise(overall = sd(x, na.rm = TRUE), within = sd(x - m, na.rm = TRUE))
dplyr 		Overall and within standard deviations
The between part is the standard deviation of the id means: summarise(data, m = mean(x, na.rm = TRUE), .by = id) then sd(m).
Stata reference · R reference

Survival analysis

Stata R Purpose, qualifications and sources 
estat concordance
 
survival::concordance(fit)
survival 		Censored-outcome concordance after Cox regression
Differs from Stata: Fixed workflow; adapt the named data columns explicitly. Censoring weights and handling of tied risks/times affect the statistic.
Stata reference · R reference 
estat phtest, detail
 
survival::cox.zph(fit, transform = "identity")
survival 		Cox proportional-hazards diagnostic
Differs from Stata: Fixed workflow; adapt the named data columns explicitly. Schoenfeld-residual implementation and default time transformations differ.
Stata reference · R reference 
ltable time event, intervals(0 2 4 6)
 
{ ends <- c(2, 4, 6); starts <- c(0, 2, 4); n <- vapply(starts, function(t) sum(data$time >= t), integer(1)); d <- vapply(seq_along(ends), function(i) sum(data$time >= starts[i] & data$time < ends[i] & data$event == 1), integer(1)); w <- vapply(seq_along(ends), function(i) sum(data$time >= starts[i] & data$time < ends[i] & data$event == 0), integer(1)); q <- d / (n - w / 2); data.frame(start = starts, end = ends, at_risk = n, failures = d, withdrawals = w, survival = cumprod(1 - q)) }
base 		Actuarial life-table survival over fixed intervals
Differs from Stata: Fixed workflow; adapt the named data columns explicitly. Uses the half-withdrawal actuarial adjustment and left-closed intervals; censoring is assumed uniform within each interval. Boundary conventions and intervals must match.
Stata reference · R reference 
predict baseS, basesurv
 
{ h <- survival::basehaz(fit, centered = FALSE); data.frame(time = h$time, survival = exp(-h$hazard)) }
survival 		Baseline Cox survival from cumulative hazard
Fixed workflow; adapt the named data columns explicitly. Baseline refers to zero covariates; verify the intended reference profile. Stata basesurv uses a product-limit construction, so exp(-cumulative hazard) is an alternative estimate rather than exact numerical parity.
Stata reference · R reference 
predict basehaz, basechazard
 
survival::basehaz(fit, centered = FALSE)
survival 		Baseline cumulative hazard at zero covariates
Differs from Stata: Fixed workflow; adapt the named data columns explicitly. Cox baseline is centered at zero explicitly; Stata stores results on observation rows whereas R returns unique event/censoring times.
Stata reference · R reference 
predict median_time, median
 
stats::predict(fit, type = "quantile", p = .5)
survival 		Conditional median survival after loglogistic AFT
Differs from Stata: Fixed workflow; adapt the named data columns explicitly. Distribution-specific predicted survival quantiles; not the mean time.
Stata reference · R reference 
predict residual*, deviance
 
stats::residuals(fit, type = "deviance")
survival 		Deviance residuals after Cox regression
Differs from Stata: Fixed workflow; adapt the named data columns explicitly. Schoenfeld residuals have one row per event; score/influence residuals have one row per observation.
Stata reference · R reference 
predict residual*, dfbeta
 
stats::residuals(fit, type = "dfbeta")
survival 		Cox coefficient influence residuals
Differs from Stata: Fixed workflow; adapt the named data columns explicitly. Schoenfeld residuals have one row per event; score/influence residuals have one row per observation.
Stata reference · R reference 
predict residual*, mgale
 
stats::residuals(fit, type = "martingale")
survival 		Martingale residuals after Cox regression
Differs from Stata: Fixed workflow; adapt the named data columns explicitly. Schoenfeld residuals have one row per event; score/influence residuals have one row per observation.
Stata reference · R reference 
predict residual*, scaledsch
 
stats::residuals(fit, type = "scaledsch")
survival 		Scaled Schoenfeld residuals
Differs from Stata: Fixed workflow; adapt the named data columns explicitly. Schoenfeld residuals have one row per event; score/influence residuals have one row per observation.
Stata reference · R reference 
predict residual*, schoenfeld
 
stats::residuals(fit, type = "schoenfeld")
survival 		Event-specific Schoenfeld residuals
Differs from Stata: Fixed workflow; adapt the named data columns explicitly. Schoenfeld residuals have one row per event; score/influence residuals have one row per observation.
Stata reference · R reference 
predict residual*, scores
 
stats::residuals(fit, type = "score")
survival 		Cox score contributions
Differs from Stata: Fixed workflow; adapt the named data columns explicitly. Schoenfeld residuals have one row per event; score/influence residuals have one row per observation.
Stata reference · R reference 
predict value, csnell
 
stats::predict(fit, type = "expected", reference = "zero")
survival 		Cox-Snell expected-event quantities
Differs from Stata: Fixed workflow; adapt the named data columns explicitly. Centering is explicitly zero for relative predictions; expected-event quantities depend on each observed follow-up time.
Stata reference · R reference 
predict value, hr
 
stats::predict(fit, type = "risk", reference = "zero")
survival 		Cox relative-hazard scores
Differs from Stata: Fixed workflow; adapt the named data columns explicitly. Centering is explicitly zero for relative predictions; expected-event quantities depend on each observed follow-up time.
Stata reference · R reference 
predict value, xb
 
stats::predict(fit, type = "lp", reference = "zero")
survival 		Cox linear predictors
Differs from Stata: Fixed workflow; adapt the named data columns explicitly. Centering is explicitly zero for relative predictions; expected-event quantities depend on each observed follow-up time.
Stata reference · R reference 
snapspan id time event, generate(start)
 
{ out <- data[order(data$id, data$time), ]; previous <- ave(out$time, out$id, FUN = function(v) c(NA_real_, head(v, -1))); previous_x <- ave(out$x, out$id, FUN = function(v) c(NA_real_, head(v, -1))); out$start <- previous; out$x <- previous_x; out[!is.na(out$start), ] }
stats 		Convert snapshots to intervals with covariates applying forward
Fixed workflow; adapt the named data columns explicitly. The first snapshot defines interval entry, event indicators apply at interval exit, and ordinary covariates apply forward. This explicit workflow drops initial zero-duration rows; check Stata's retained baseline rows before matching row counts.
Stata reference · R reference 
stbase
 
data[order(data$id, data$start), ][!duplicated(data[order(data$id, data$start), "id"]), ]
base 		Keep baseline record for each subject
Fixed workflow; adapt the named data columns explicitly. Record-selection workflow only; stbase may also transform survival-time fields.
Stata reference · R reference 
stci, median
 
stats::quantile(survival::survfit(s ~ 1), probs = .5)
survival 		Median survival with confidence limits
Differs from Stata: Fixed workflow; adapt the named data columns explicitly. Median interpolation and confidence-limit conventions can differ.
Stata reference · R reference 
stci, rmean
 
summary(survival::survfit(s ~ 1), rmean = "common")$table
survival 		Restricted mean survival through time nine
Differs from Stata: rmean="common" restricts the survival integral at the largest observed follow-up time, as Stata stci, rmean does; it does not extrapolate beyond follow-up. Match the analysis sample and censoring declaration.
Stata reference · R reference 
stcox x1 x2
 
survival::coxph(survival::Surv(time, died) ~ x1 + x2, data = data)
survival 		Cox proportional hazards model
Differs from Stata: exp(coef(fit)) gives the hazard ratios that stcox reports. coxph() defaults to Efron handling of ties; Stata stcox defaults to Breslow. Set ties = "breslow" to match tied-event likelihoods.
Stata reference · R reference 
stcox c.x##i.group
 
survival::coxph(survival::Surv(time, event) ~ x * factor(group), ties = "breslow", data = data)
survival 		Cox treatment-covariate interaction
Differs from Stata: Fixed workflow; adapt the named data columns explicitly. Match ties, risk intervals, weights and reference inference explicitly.
Stata reference · R reference 
stcox x z
 
survival::coxph(survival::Surv(time, event) ~ x + z, data = data, ties = "breslow")
survival 		Cox regression with Breslow ties
Differs from Stata: Fixed workflow; adapt the named data columns explicitly. Match ties, risk intervals, weights and reference inference explicitly.
Stata reference · R reference 
stcox x z [pw=w], vce(robust)
 
survival::coxph(survival::Surv(time, event) ~ x + z, ties = "breslow", data = data, weights = w, robust = TRUE)
survival 		Weighted Cox model with sandwich covariance
Differs from Stata: Fixed workflow; adapt the named data columns explicitly. Match ties, risk intervals, weights and reference inference explicitly.
Stata reference · R reference 
stcox x z, efron
 
survival::coxph(survival::Surv(time, event) ~ x + z, data = data, ties = "efron")
survival 		Cox regression with Efron ties
Differs from Stata: Fixed workflow; adapt the named data columns explicitly. Match ties, risk intervals, weights and reference inference explicitly.
Stata reference · R reference 
stcox x z, exactp
 
survival::coxph(survival::Surv(time, event) ~ x + z, data = data, ties = "exact")
survival 		Cox regression with exact partial likelihood
Differs from Stata: Fixed workflow; adapt the named data columns explicitly. Match ties, risk intervals, weights and reference inference explicitly.
Stata reference · R reference 
stcox x z, shared(id)
 
survival::coxph(survival::Surv(time, event) ~ x + z + frailty(id, distribution = "gamma"), ties = "breslow", data = data)
survival 		Shared gamma frailty Cox model
Differs from Stata: Fixed workflow; adapt the named data columns explicitly. Match ties, risk intervals, weights and reference inference explicitly.
Stata reference · R reference 
stcox x z, strata(group)
 
survival::coxph(survival::Surv(time, event) ~ x + z + strata(group), ties = "breslow", data = data)
survival 		Stratified Cox baseline hazards
Differs from Stata: Fixed workflow; adapt the named data columns explicitly. Match ties, risk intervals, weights and reference inference explicitly.
Stata reference · R reference 
stcox x z, vce(cluster id)
 
survival::coxph(survival::Surv(time, event) ~ x + z, ties = "breslow", data = data, cluster = id)
survival 		Cluster-robust Cox inference
Differs from Stata: Fixed workflow; adapt the named data columns explicitly. Match ties, risk intervals, weights and reference inference explicitly.
Stata reference · R reference 
stcox x, tvc(z) texp(ln(_t))
 
survival::coxph(survival::Surv(time, event) ~ x + tt(z), ties = "breslow", data = data, tt = function(z, t, ...) z * log(t))
survival 		Cox time-varying log-time coefficient
Differs from Stata: Fixed workflow; adapt the named data columns explicitly. Match ties, risk intervals, weights and reference inference explicitly.
Stata reference · R reference 
stcrreg x z, compete(cause==2)
 
{ expanded <- survival::finegray(survival::Surv(time, cause) ~ x + z + id, data = data, etype = "target"); survival::coxph(survival::Surv(fgstart, fgstop, fgstatus) ~ x + z, data = expanded, weights = fgwt, cluster = id) }
survival 		Fine-Gray subdistribution-hazard workflow
Differs from Stata: Fixed workflow; adapt the named data columns explicitly. Weighted expanded-data implementation; censoring weights and variance choices can differ.
Stata reference · R reference 
stdescribe
 
data.frame(subjects = nrow(data), failures = sum(data$event), time_at_risk = sum(data$time))
base 		Describe single-record survival data
Fixed workflow; adapt the named data columns explicitly. Single-record, zero-entry fixture; multiple-record subjects require distinct-ID counting.
Stata reference · R reference 
stfill z, forward
 
{ out <- data[order(data$id, data$start), ]; out$z <- ave(out$z, out$id, FUN = function(v) zoo::na.locf(v, na.rm = FALSE)); out }
zoo 		Carry observed covariates forward within survival subjects
Differs from Stata: Fixed workflow; adapt the named data columns explicitly. Requires correctly ordered non-overlapping spells. Leading missing values remain missing; later information is never carried backward.
Stata reference · R reference 
stgen ever_x = ever(x==1)
 
transform(data, ever_x = ave(x == 1, id, FUN = function(v) as.integer(any(v))))
stats 		Mark subjects ever exposed during their observed survival history
Differs from Stata: Fixed workflow; adapt the named data columns explicitly. All observed spells contribute to the history. Missing exposure values require an explicit convention rather than treating NA as false.
Stata reference · R reference 
stintreg x, interval(lower upper) distribution(lognormal)
 
survival::survreg(survival::Surv(lower, upper, type = "interval2") ~ x, data = data, dist = "lognormal")
survival 		Lognormal regression for interval-censored outcomes
Differs from Stata: Stata requires interval(lower upper). The Weibull time option selects AFT coefficients, matching survreg rather than Stata default PH coefficients. For interval2, equal endpoints are exact events, NA lower endpoints denote left censoring, and NA upper endpoints right censoring; encode Stata left-censoring bounds explicitly.
Stata reference · R reference 
stintreg x, interval(lower upper) distribution(weibull) time
 
survival::survreg(survival::Surv(lower, upper, type = "interval2") ~ x, data = data, dist = "weibull")
survival 		Weibull regression for interval-censored outcomes
Differs from Stata: Stata requires interval(lower upper). The Weibull time option selects AFT coefficients, matching survreg rather than Stata default PH coefficients. For interval2, equal endpoints are exact events, NA lower endpoints denote left censoring, and NA upper endpoints right censoring; encode Stata left-censoring bounds explicitly.
Stata reference · R reference 
stir group
 
{ counts <- tapply(data$event, data$group, sum); times <- tapply(data$time, data$group, sum); stats::poisson.test(counts, T = times) }
stats 		Compare two incidence rates
Differs from Stata: Fixed workflow; adapt the named data columns explicitly. R gives a conditional exact test and interval; Stata default intervals may use a different method.
Stata reference · R reference 
stjoin
 
{ out <- stats::aggregate(cbind(time = time - start, event) ~ id, data = spells, sum); names(out)[2] <- "duration"; out }
stats 		Rejoin adjacent constant-covariate spells
Fixed workflow; adapt the named data columns explicitly. Only valid when adjacent spells have identical covariates and no intervening event; inspect before collapsing.
Stata reference · R reference 
stmh exposed stratum
 
with(data, sum(events1 * time0 / (time1 + time0)) / sum(events0 * time1 / (time1 + time0)))
base 		Mantel-Haenszel stratified incidence-rate-ratio point estimate
Fixed workflow; adapt the named data columns explicitly. Input is pre-aggregated failure counts and person-time by exposure/stratum. This supplies the rate-ratio point estimate only, not Stata's variance or trend tests.
Stata reference · R reference 
stptime
 
{ pt <- sum(data$time); events <- sum(data$event); c(person_time = pt, failures = events, rate = events / pt) }
base 		Crude incidence rate from person-time
Fixed workflow; adapt the named data columns explicitly. Assumes zero entry times and non-overlapping observation spells.
Stata reference · R reference 
strate group, per(1000)
 
{ out <- stats::aggregate(cbind(time, event) ~ group, data, sum); out$rate_per_1000 <- 1000 * out$event / out$time; out }
stats 		Group-specific rates per thousand person-time units
Fixed workflow; adapt the named data columns explicitly. Point-rate workflow; confidence intervals need a separate choice.
Stata reference · R reference 
streg x, distribution(weibull)
 
survival::survreg(survival::Surv(time, died) ~ x, data = data, dist = "weibull")
survival 		Parametric survival model
Differs from Stata: survreg() reports the accelerated failure-time form; Stata's Weibull default is the hazard form, so signs and scales differ.
Stata reference · R reference 
streg x z, distribution(exponential) time
 
survival::survreg(survival::Surv(time, event) ~ x + z, data = data, dist = "exponential")
survival 		exponential accelerated-failure-time model
Differs from Stata: Fixed workflow; adapt the named data columns explicitly. survreg reports AFT coefficients; Weibull PH coefficients require sign and scale conversion.
Stata reference · R reference 
streg x z, distribution(loglogistic) time
 
survival::survreg(survival::Surv(time, event) ~ x + z, data = data, dist = "loglogistic")
survival 		loglogistic accelerated-failure-time model
Differs from Stata: Fixed workflow; adapt the named data columns explicitly. survreg reports AFT coefficients; Weibull PH coefficients require sign and scale conversion.
Stata reference · R reference 
streg x z, distribution(lognormal) time
 
survival::survreg(survival::Surv(time, event) ~ x + z, data = data, dist = "lognormal")
survival 		lognormal accelerated-failure-time model
Differs from Stata: Fixed workflow; adapt the named data columns explicitly. survreg reports AFT coefficients; Weibull PH coefficients require sign and scale conversion.
Stata reference · R reference 
streg x z, distribution(weibull) time
 
survival::survreg(survival::Surv(time, event) ~ x + z, data = data, dist = "weibull")
survival 		weibull accelerated-failure-time model
Differs from Stata: Fixed workflow; adapt the named data columns explicitly. survreg reports AFT coefficients; Weibull PH coefficients require sign and scale conversion.
Stata reference · R reference 
streg x, distribution(loglogistic) strata(group) time
 
survival::survreg(survival::Surv(time, event) ~ x + factor(group) + strata(group), data = data, dist = "loglogistic")
survival 		loglogistic AFT model with group-specific scale
Differs from Stata: factor(group) gives stratum-specific intercepts and strata(group) gives stratum-specific scales, as in Stata streg, strata(). survreg reports AFT coefficients; convert Weibull/exponential coefficients before comparing Stata hazard-ratio output.
Stata reference · R reference 
streg x, distribution(lognormal) strata(group) time
 
survival::survreg(survival::Surv(time, event) ~ x + factor(group) + strata(group), data = data, dist = "lognormal")
survival 		lognormal AFT model with group-specific scale
Differs from Stata: factor(group) gives stratum-specific intercepts and strata(group) gives stratum-specific scales, as in Stata streg, strata(). survreg reports AFT coefficients; convert Weibull/exponential coefficients before comparing Stata hazard-ratio output.
Stata reference · R reference 
streg x, distribution(weibull) strata(group) time
 
survival::survreg(survival::Surv(time, event) ~ x + factor(group) + strata(group), data = data, dist = "weibull")
survival 		weibull AFT model with group-specific scale
Differs from Stata: factor(group) gives stratum-specific intercepts and strata(group) gives stratum-specific scales, as in Stata streg, strata(). survreg reports AFT coefficients; convert Weibull/exponential coefficients before comparing Stata hazard-ratio output.
Stata reference · R reference 
sts graph
 
plot(survival::survfit(survival::Surv(time, died) ~ 1, data = data))
survival 		Kaplan-Meier curve
Stata reference · R reference 
sts test group
 
survival::survdiff(survival::Surv(time, died) ~ group, data = data)
survival 		Log-rank test
Stata reference · R reference 
sts list
 
survival::survfit(s ~ 1)
survival 		Kaplan-Meier survival curve data
Differs from Stata: Fixed workflow; adapt the named data columns explicitly. R returns curve data; confidence transformations and tied-event handling should be specified for comparisons.
Stata reference · R reference 
sts list, by(group)
 
survival::survfit(s ~ data$group)
survival 		Group-specific Kaplan-Meier curves
Differs from Stata: Fixed workflow; adapt the named data columns explicitly. R returns curve data; confidence transformations and tied-event handling should be specified for comparisons.
Stata reference · R reference 
sts list, cumhaz
 
{ f <- survival::survfit(s ~ 1); data.frame(time = f$time, cumhaz = f$cumhaz) }
survival 		Nelson-Aalen cumulative hazard
Differs from Stata: Fixed workflow; adapt the named data columns explicitly. R returns curve data; confidence transformations and tied-event handling should be specified for comparisons.
Stata reference · R reference 
sts list, failure
 
{ f <- survival::survfit(s ~ 1); data.frame(time = f$time, failure = 1 - f$surv) }
survival 		Cumulative failure probability
Differs from Stata: Fixed workflow; adapt the named data columns explicitly. R returns curve data; confidence transformations and tied-event handling should be specified for comparisons.
Stata reference · R reference 
sts list, level(95)
 
survival::survfit(s ~ 1, conf.type = "plain", conf.int = .95)
survival 		Kaplan-Meier interval with plain transformation
Fixed workflow; adapt the named data columns explicitly. Explicit R confidence transformation; Stata's displayed default may differ.
Stata reference · R reference 
sts list, level(95)
 
survival::survfit(s ~ 1, conf.type = "log", conf.int = .95)
survival 		Kaplan-Meier interval with log transformation
Fixed workflow; adapt the named data columns explicitly. Explicit R confidence transformation; Stata's displayed default may differ.
Stata reference · R reference 
sts list, level(95)
 
survival::survfit(s ~ 1, conf.type = "log-log", conf.int = .95)
survival 		Kaplan-Meier interval with log-log transformation
Fixed workflow; adapt the named data columns explicitly. Explicit R confidence transformation; Stata's displayed default may differ.
Stata reference · R reference 
sts test group
 
survival::survdiff(s ~ data$group, rho = 0)
survival 		Log-rank comparison of survival curves
Differs from Stata: Fixed workflow; adapt the named data columns explicitly. Unstratified two-group log-rank test.
Stata reference · R reference 
stset time, failure(died)
 
survival::Surv(data$time, data$died)
survival 		Declare survival-time data
R has no declaration step: Surv(time, died) goes on the left of each model formula.
Stata reference · R reference 
stset exit, enter(time entry) failure(event)
 
survival::Surv(data$entry, data$time, data$event)
survival 		Delayed-entry counting-process response
Differs from Stata: Fixed workflow; adapt the named data columns explicitly. Supply entry and exit explicitly; multiple spells require one row per risk interval.
Stata reference · R reference 
stset time, failure(event)
 
survival::Surv(data$time, data$event)
survival 		Right-censored survival response
Differs from Stata: Fixed workflow; adapt the named data columns explicitly. Creates a response object rather than declaring global dataset state.
Stata reference · R reference 
stsplit period, at(3 6)
 
{ Surv <- survival::Surv; survival::survSplit(Surv(time, event) ~ ., data = data, cut = c(3, 6), start = "start", end = "time", event = "event", episode = "period") }
survival 		Split risk intervals at fixed cut points
Differs from Stata: Fixed workflow; adapt the named data columns explicitly. Returns expanded interval data; the episode variable is an interval index.
Stata reference · R reference 
stsum
 
survival::survfit(survival::Surv(time, died) ~ 1, data = data)
survival 		Summary of survival times
Stata reference · R reference 
stsum, by(group)
 
stats::aggregate(cbind(person_time = time, failures = event) ~ group, data = data, FUN = sum)
stats 		Person-time and failure counts by group
Fixed workflow; adapt the named data columns explicitly. Subset of stsum output; no survival quantiles are computed.
Stata reference · R reference 
sttocc, number(2)
 
{ cases <- which(data$event == 1); sets <- lapply(seq_along(cases), function(j) { i <- cases[j]; at_risk <- which(data$entry <= data$time[i] & data$time >= data$time[i] & data$id != data$id[i]); controls <- sample(at_risk, min(2L, length(at_risk)), replace = FALSE); out <- data[c(i, controls), ]; out$case <- c(1L, rep(0L, length(controls))); out$set <- j; out$failure_time <- data$time[i]; out }); do.call(rbind, sets) }
base 		Sample nested case-control risk sets from a cohort
Fixed workflow; adapt the named data columns explicitly. One-record cohort with no tied failures. Controls are sampled from those at risk for each failure and may recur across sets; tied-failure random ordering and multiple spells need a fuller implementation.
Stata reference · R reference 
stvary x z
 
lapply(split(data, data$id), function(d) vapply(d[c("x", "z")], function(v) c(varies = length(unique(v[!is.na(v)])) > 1, any_missing = anyNA(v)), logical(2)))
base 		Check within-subject covariate variation and missingness
Fixed workflow; adapt the named data columns explicitly. Returns per-subject flags rather than Stata's aggregate summary table; ignores sampling weights as stvary does.
Stata reference · R reference

Survey data

Stata R Purpose, qualifications and sources 
svy: logit y x
 
survey::svyglm(y ~ x, design = design, family = quasibinomial())
survey 		Survey-weighted logit
Differs from Stata: Match design and missing-data exclusions first; singleton-PSU policy, degrees of freedom and confidence-interval transformations can differ between R and Stata.
Stata reference · R reference 
svy: mean x
 
summarise(design, x = srvyr::survey_mean(x, na.rm = TRUE))
srvyr 		Survey-weighted mean
Differs from Stata: Match design and missing-data exclusions first; singleton-PSU policy, degrees of freedom and confidence-interval transformations can differ between R and Stata.
Stata reference · R reference 
svy: regress y x
 
survey::svyglm(y ~ x, design = design)
survey 		Survey-weighted regression
Differs from Stata: Match design and missing-data exclusions first; singleton-PSU policy, degrees of freedom and confidence-interval transformations can differ between R and Stata.
Stata reference · R reference 
svy: tabulate x
 
design |> group_by(x) |> summarise(p = srvyr::survey_prop())
srvyr 		Survey-weighted proportions
Differs from Stata: Match design and missing-data exclusions first; singleton-PSU policy, degrees of freedom and confidence-interval transformations can differ between R and Stata.
Stata reference · R reference 
svy, subpop(rural): mean x
 
design |> filter(rural == 1) |> summarise(x = srvyr::survey_mean(x, na.rm = TRUE))
srvyr 		Subpopulation estimate
Differs from Stata: Filtering a survey design keeps the full design for the variance, as subpop() does. Match design and missing-data exclusions first; singleton-PSU policy, degrees of freedom and confidence-interval transformations can differ between R and Stata.
Stata reference · R reference 
svy: mean x, over(g)
 
design |> group_by(g) |> summarise(x = srvyr::survey_mean(x, na.rm = TRUE))
srvyr 		Survey-weighted means by group
Differs from Stata: Match design and missing-data exclusions first; singleton-PSU policy, degrees of freedom and confidence-interval transformations can differ between R and Stata.
Stata reference · R reference 
svy: proportion x
 
design |> group_by(x) |> summarise(p = srvyr::survey_prop(vartype = "ci"))
srvyr 		Survey-weighted proportions with intervals
Differs from Stata: Match design and missing-data exclusions first; singleton-PSU policy, degrees of freedom and confidence-interval transformations can differ between R and Stata.
Stata reference · R reference 
svydescribe
 
summary(design)
base 		Describe the survey design
Stata reference · R reference 
svyset psu [pweight = wt], strata(stratum)
 
design <- srvyr::as_survey_design(data, ids = psu, strata = stratum, weights = wt)
srvyr 		Declare a survey design
Differs from Stata: Match design and missing-data exclusions first; singleton-PSU policy, degrees of freedom and confidence-interval transformations can differ between R and Stata.
Stata reference · R reference

Graphs

Stata R Purpose, qualifications and sources 
binscatter y x
 
binsreg::binsreg(y, x, data = data)
binsreg 		Binned scatter plot
Differs from Stata: statar::stat_binmean() adds binned means to a ggplot: ggplot(data, aes(x, y)) + statar::stat_binmean(n = 20). binsreg() defaults to data-driven bin selection and inference; binscatter defaults to fixed quantile bins and an OLS fit. Match bins and fit settings explicitly.
Stata reference · R reference 
coefplot
 
modelsummary::modelplot(fit)
modelsummary 		Coefficient plot
Stata reference · R reference 
dotplot y, over(group)
 
ggplot2::ggplot(data, ggplot2::aes(group, y)) + ggplot2::geom_point(position = ggplot2::position_jitter(width = 0.15, height = 0))
ggplot2 		Show individual observations by category
Differs from Stata: This uses horizontally jittered observations rather than Stata dot-stacking rules.
Stata reference · R reference 
graph bar (count), over(x)
 
ggplot(data, aes(x)) + geom_bar()
ggplot2 		Bar chart of counts
Stata reference · R reference 
graph bar (mean) y, over(group)
 
ggplot(data, aes(group, y)) + stat_summary(fun = mean, geom = "bar")
ggplot2 		Bar chart of group means
Stata reference · R reference 
graph box y, over(group)
 
ggplot(data, aes(group, y)) + geom_boxplot()
ggplot2 		Box plot
Differs from Stata: ggplot2 boxplot hinges use R quantile rules; Stata uses Stata percentiles, which may differ in small samples and with ties.
Stata reference · R reference 
graph export figure.png
 
ggsave("figure.png")
ggplot2 		Save the last graph
Stata reference · R reference 
graph pie, over(group)
 
ggplot(data, aes(x = "", fill = group)) + geom_bar() + coord_polar(theta = "y")
ggplot2 		Pie chart
Stata reference · R reference 
graph combine g1.gph g2.gph
 
patchwork::wrap_plots(p1, p2)
patchwork 		Combine plots
With patchwork attached, p1 + p2 does the same.
Stata reference · R reference 
graph dot (mean) y, over(group)
 
ggplot(data, aes(y, group)) + stat_summary(fun = mean, geom = "point")
ggplot2 		Dot chart
Stata reference · R reference 
graph hbar (mean) y, over(group)
 
ggplot(data, aes(group, y)) + stat_summary(fun = mean, geom = "bar") + coord_flip()
ggplot2 		Horizontal bar chart
Stata reference · R reference 
graph hbox y, over(group)
 
ggplot(data, aes(y, group)) + geom_boxplot()
ggplot2 		Horizontal box plot
Differs from Stata: ggplot2 boxplot hinges use R quantile rules; Stata uses Stata percentiles, which may differ in small samples and with ties.
Stata reference · R reference 
graph matrix x1 x2 x3
 
pairs(select(data, x1, x2, x3))
graphics 		Scatter-plot matrix
GGally::ggpairs() draws a ggplot2 version.
Stata reference · R reference 
graph save fig.gph
 
saveRDS(p, "fig.rds")
base 		Save a plot to edit later
Stata reference · R reference 
graph bar (count), over(group)
 
ggplot2::ggplot(data, ggplot2::aes(group)) + ggplot2::geom_bar()
ggplot2 		Plot group frequencies
Differs from Stata: Reproduces the displayed count, percent or summary layer, rather than Stata graph display defaults. Match category order and missingness; ggplot themes, axis scales and bar annotations follow R conventions.
Stata reference · R reference 
graph bar (mean) y, over(group)
 
ggplot2::ggplot(data, ggplot2::aes(group, y)) + ggplot2::stat_summary(fun = mean, geom = "col")
ggplot2 		Plot group means as bars
Differs from Stata: Reproduces the displayed count, percent or summary layer, rather than Stata graph display defaults. Match category order and missingness; ggplot themes, axis scales and bar annotations follow R conventions.
Stata reference · R reference 
graph bar (mean) y, over(group) blabel(bar)
 
{ plot_data <- aggregate(y ~ group, data, mean); ggplot2::ggplot(plot_data, ggplot2::aes(group, y)) + ggplot2::geom_col() + ggplot2::geom_text(ggplot2::aes(label = round(y, 1)), vjust = -0.3) }
ggplot2 		Label summary bars with their values
Differs from Stata: Reproduces the displayed count, percent or summary layer, rather than Stata graph display defaults. Match category order and missingness; ggplot themes, axis scales and bar annotations follow R conventions.
Stata reference · R reference 
graph bar (median) y, over(group)
 
ggplot2::ggplot(data, ggplot2::aes(group, y)) + ggplot2::stat_summary(fun = median, geom = "col")
ggplot2 		Plot group medians as bars
Differs from Stata: Reproduces the displayed count, percent or summary layer, rather than Stata graph display defaults. Match category order and missingness; ggplot themes, axis scales and bar annotations follow R conventions.
Stata reference · R reference 
graph bar (percent), over(group)
 
{ counts <- as.data.frame(table(data$group)); ggplot2::ggplot(counts, ggplot2::aes(Var1, Freq / sum(Freq) * 100)) + ggplot2::geom_col() }
ggplot2 		Plot group percentages
Differs from Stata: Reproduces the displayed count, percent or summary layer, rather than Stata graph display defaults. Match category order and missingness; ggplot themes, axis scales and bar annotations follow R conventions.
Stata reference · R reference 
graph bar (sum) y, over(group)
 
ggplot2::ggplot(data, ggplot2::aes(group, y)) + ggplot2::stat_summary(fun = sum, geom = "col")
ggplot2 		Plot group totals as bars
Differs from Stata: Reproduces the displayed count, percent or summary layer, rather than Stata graph display defaults. Match category order and missingness; ggplot themes, axis scales and bar annotations follow R conventions.
Stata reference · R reference 
graph box y, over(group)
 
ggplot2::ggplot(data, ggplot2::aes(group, y)) + ggplot2::geom_boxplot()
ggplot2 		Compare distributions with box plots
Differs from Stata: Box-plot hinges and whisker defaults should be checked when numerical parity is required.
Stata reference · R reference 
graph box y, over(group) nooutsides
 
ggplot2::ggplot(data, ggplot2::aes(group, y)) + ggplot2::geom_boxplot(outlier.shape = NA)
ggplot2 		Hide box-plot outlier markers
Differs from Stata: Box-plot hinges and whisker defaults should be checked when numerical parity is required.
Stata reference · R reference 
graph dot (mean) y, over(group)
 
ggplot2::ggplot(data, ggplot2::aes(y, group)) + ggplot2::stat_summary(fun = mean, geom = "point")
ggplot2 		Plot group means as dots
Differs from Stata: Reproduces the displayed count, percent or summary layer, rather than Stata graph display defaults. Match category order and missingness; ggplot themes, axis scales and bar annotations follow R conventions.
Stata reference · R reference 
graph hbox y, over(group)
 
ggplot2::ggplot(data, ggplot2::aes(y, group)) + ggplot2::geom_boxplot()
ggplot2 		Draw horizontal box plots
Differs from Stata: Box-plot hinges and whisker defaults should be checked when numerical parity is required.
Stata reference · R reference 
heatplot z y x
 
ggplot2::ggplot(data, ggplot2::aes(x, y, z = z)) + ggplot2::stat_summary_2d(fun = mean)
ggplot2 		Heat map of the mean of z over binned x and y
Differs from Stata: Community command. Both average z within cells, but bin counts and boundaries differ; set bins explicitly (bins() in Stata, bins = in R) when comparing. Colour ramps follow each program's defaults.
Stata reference · R reference 
heatplot z i.y i.x
 
ggplot2::ggplot(stats::aggregate(z ~ x + y, data = data, FUN = mean), ggplot2::aes(factor(x), factor(y), fill = z)) + ggplot2::geom_tile()
ggplot2 		Heat map of the mean of z over categories of x and y
Differs from Stata: Community command. One tile per observed x-y combination, filled with the mean of z; empty combinations are left blank in both.
Stata reference · R reference 
histogram x
 
ggplot(data, aes(x)) + geom_histogram()
ggplot2 		Histogram
Differs from Stata: geom_histogram() defaults to counts and 30 bins; Stata histogram defaults to density with a different bin rule. Specify breaks/binwidth and the y statistic.
Stata reference · R reference 
histogram y, by(group)
 
ggplot2::ggplot(data, ggplot2::aes(y)) + ggplot2::geom_histogram(bins = 10) + ggplot2::facet_wrap(~group)
ggplot2 		Compare histograms across groups
Differs from Stata: Set boundaries and interval closure explicitly when matching Stata bin counts.
Stata reference · R reference 
histogram y, density
 
ggplot2::ggplot(data, ggplot2::aes(y, ggplot2::after_stat(density))) + ggplot2::geom_histogram(bins = 10)
ggplot2 		Histogram on the density scale
Differs from Stata: Set boundaries and interval closure explicitly when matching Stata bin counts.
Stata reference · R reference 
histogram y, frequency
 
ggplot2::ggplot(data, ggplot2::aes(y)) + ggplot2::geom_histogram(bins = 10)
ggplot2 		Histogram on the frequency scale
Differs from Stata: Set boundaries and interval closure explicitly when matching Stata bin counts.
Stata reference · R reference 
histogram y, percent
 
ggplot2::ggplot(data, ggplot2::aes(y, ggplot2::after_stat(count / sum(count) * 100))) + ggplot2::geom_histogram(bins = 10)
ggplot2 		Histogram on the percentage scale
Differs from Stata: Set boundaries and interval closure explicitly when matching Stata bin counts.
Stata reference · R reference 
histogram y, width(2) start(0)
 
ggplot2::ggplot(data, ggplot2::aes(y)) + ggplot2::geom_histogram(binwidth = 2, boundary = 0)
ggplot2 		Choose histogram width and boundary
Differs from Stata: Set boundaries and interval closure explicitly when matching Stata bin counts.
Stata reference · R reference 
histogram z, discrete
 
ggplot2::ggplot(data, ggplot2::aes(z)) + ggplot2::geom_histogram(binwidth = 1, boundary = 0.5)
ggplot2 		Histogram for integer-valued observations
Differs from Stata: Set boundaries and interval closure explicitly when matching Stata bin counts.
Stata reference · R reference 
kdensity x
 
ggplot(data, aes(x)) + geom_density()
ggplot2 		Density plot
Differs from Stata: Default kernel-density bandwidth rules differ; specify kernel, bandwidth and evaluation range before comparing curves.
Stata reference · R reference 
kdensity y
 
ggplot2::ggplot(data, ggplot2::aes(y)) + ggplot2::geom_density()
ggplot2 		Plot a kernel density estimate
Differs from Stata: R and Stata bandwidth rules and local smoothing algorithms can differ; this is a graphical alternative, not an exact numerical replication.
Stata reference · R reference 
kdensity y if group == "A"
 
ggplot2::ggplot(subset(data, group == "A"), ggplot2::aes(y)) + ggplot2::geom_density()
ggplot2 		Estimate a density for a subgroup
Differs from Stata: R and Stata bandwidth rules and local smoothing algorithms can differ; this is a graphical alternative, not an exact numerical replication.
Stata reference · R reference 
kdensity y, bwidth(1)
 
ggplot2::ggplot(data, ggplot2::aes(y)) + ggplot2::geom_density(bw = 1)
ggplot2 		Specify a kernel bandwidth
Differs from Stata: R and Stata bandwidth rules and local smoothing algorithms can differ; this is a graphical alternative, not an exact numerical replication.
Stata reference · R reference 
kdensity y, kernel(gaussian)
 
ggplot2::ggplot(data, ggplot2::aes(y)) + ggplot2::geom_density(kernel = "gaussian")
ggplot2 		Use a Gaussian density kernel
Differs from Stata: R and Stata bandwidth rules and local smoothing algorithms can differ; this is a graphical alternative, not an exact numerical replication.
Stata reference · R reference 
line y x
 
ggplot(data, aes(x, y)) + geom_line()
ggplot2 		Line plot
Differs from Stata: geom_line() sorts by x within groups; Stata line/connected uses observation order without the sort option.
Stata reference · R reference 
line y x, sort
 
ggplot2::ggplot(data, ggplot2::aes(x, y)) + ggplot2::geom_line()
ggplot2 		Line plot sorted by the horizontal variable
Differs from Stata: Maps the stated axes and layer; line sorts by x in ggplot while Stata uses sort here. Plot themes, axis expansion, missing-point warnings and facet scales follow R defaults.
Stata reference · R reference 
line y x, sort by(group)
 
ggplot2::ggplot(data, ggplot2::aes(x, y)) + ggplot2::geom_line() + ggplot2::facet_wrap(~group)
ggplot2 		Draw line plots in separate group panels
Differs from Stata: Maps the stated axes and layer; line sorts by x in ggplot while Stata uses sort here. Plot themes, axis expansion, missing-point warnings and facet scales follow R defaults.
Stata reference · R reference 
line y x, sort lcolor(red)
 
ggplot2::ggplot(data, ggplot2::aes(x, y)) + ggplot2::geom_line(colour = "red")
ggplot2 		Set a line colour
Differs from Stata: Maps the stated axes and layer; line sorts by x in ggplot while Stata uses sort here. Plot themes, axis expansion, missing-point warnings and facet scales follow R defaults.
Stata reference · R reference 
lowess y x
 
ggplot(data, aes(x, y)) + geom_point() + geom_smooth(method = "loess")
ggplot2 		Smoothed scatter plot
Differs from Stata: LOESS local polynomials differ from Stata LOWESS smoothing, robustness iterations and bandwidth defaults; the fitted curves need not coincide.
Stata reference · R reference 
lowess y x, bwidth(0.5)
 
ggplot2::ggplot(data, ggplot2::aes(x, y)) + ggplot2::geom_smooth(method = "loess", formula = y ~ x, span = 0.5, se = FALSE)
ggplot2 		Draw a local regression smooth
Differs from Stata: R and Stata bandwidth rules and local smoothing algorithms can differ; this is a graphical alternative, not an exact numerical replication.
Stata reference · R reference 
pchi y, df(2)
 
graphics::plot(ppoints(length(data$y)), pchisq(sort(data$y), df = 2))
graphics 		Compare empirical and chi-squared probabilities
Differs from Stata: A diagnostic plotting workflow; plotting-position and standardization defaults can differ from Stata.
Stata reference · R reference 
pnorm y
 
graphics::plot(ppoints(length(data$y)), pnorm(as.numeric(scale(sort(data$y)))), xlab = "Empirical", ylab = "Normal")
graphics 		Assess a standardized normal P-P relationship
Differs from Stata: A diagnostic plotting workflow; plotting-position and standardization defaults can differ from Stata.
Stata reference · R reference 
qchi y, df(2)
 
graphics::plot(qchisq(ppoints(length(data$y)), df = 2), sort(data$y))
graphics 		Compare sample quantiles with chi-squared quantiles
Differs from Stata: A diagnostic plotting workflow; plotting-position and standardization defaults can differ from Stata.
Stata reference · R reference 
qnorm x
 
ggplot(data, aes(sample = x)) + geom_qq() + geom_qq_line()
ggplot2 		Normal quantile plot
Differs from Stata: geom_qq() defaults to theoretical standard-normal quantiles; qnorm scales the normal axis by the sample mean and SD. Match axis scaling to compare points.
Stata reference · R reference 
qqplot y x
 
stats::qqplot(data$x, data$y)
stats 		Compare two empirical distributions with a Q-Q plot
Differs from Stata: A diagnostic plotting workflow; plotting-position and standardization defaults can differ from Stata.
Stata reference · R reference 
quantile y
 
graphics::plot(ppoints(sum(!is.na(data$y))), sort(data$y), xlab = "Fraction", ylab = "y")
graphics 		Plot ordered data against cumulative fractions
Differs from Stata: A diagnostic plotting workflow; plotting-position and standardization defaults can differ from Stata.
Stata reference · R reference 
scatter y x
 
ggplot(data, aes(x, y)) + geom_point()
ggplot2 		Scatter plot
Stata reference · R reference 
scatter y x, mlabel(name)
 
ggplot(data, aes(x, y, label = name)) + geom_point() + geom_text()
ggplot2 		Label the points
Stata reference · R reference 
scatter y x, xline(0)
 
ggplot(data, aes(x, y)) + geom_point() + geom_vline(xintercept = 0)
ggplot2 		Add a reference line
geom_hline(yintercept = 0) is yline().
Stata reference · R reference 
scatter y x
 
ggplot2::ggplot(data, ggplot2::aes(x, y)) + ggplot2::geom_point()
ggplot2 		Scatter plot with ggplot2
Differs from Stata: Maps the stated axes and layer; line sorts by x in ggplot while Stata uses sort here. Plot themes, axis expansion, missing-point warnings and facet scales follow R defaults.
Stata reference · R reference 
scatter y x, aspectratio(1)
 
ggplot2::ggplot(data, ggplot2::aes(x, y)) + ggplot2::geom_point() + ggplot2::theme(aspect.ratio = 1)
ggplot2 		Make the plotting panel square
Differs from Stata: The R example implements the graphical intent; size, jitter and theme scales are not numerically identical to Stata.
Stata reference · R reference 
scatter y x, by(group)
 
ggplot2::ggplot(data, ggplot2::aes(x, y)) + ggplot2::geom_point() + ggplot2::facet_wrap(~group)
ggplot2 		Separate scatter plots by group
Differs from Stata: The R example implements the graphical intent; size, jitter and theme scales are not numerically identical to Stata.
Stata reference · R reference 
scatter y x, jitter(2)
 
ggplot2::ggplot(data, ggplot2::aes(x, y)) + ggplot2::geom_jitter(width = 0.2, height = 0.2)
ggplot2 		Jitter overlapping observations
Differs from Stata: The R example implements the graphical intent; size, jitter and theme scales are not numerically identical to Stata.
Stata reference · R reference 
scatter y x, legend(off)
 
ggplot2::ggplot(data, ggplot2::aes(x, y)) + ggplot2::geom_point(ggplot2::aes(colour = group)) + ggplot2::theme(legend.position = "none")
ggplot2 		Hide a graph legend
Differs from Stata: The R example implements the graphical intent; size, jitter and theme scales are not numerically identical to Stata.
Stata reference · R reference 
scatter y x, mcolor(red)
 
ggplot2::ggplot(data, ggplot2::aes(x, y)) + ggplot2::geom_point(colour = "red")
ggplot2 		Set the marker colour
Differs from Stata: The R example implements the graphical intent; size, jitter and theme scales are not numerically identical to Stata.
Stata reference · R reference 
scatter y x, mlabel(label)
 
ggplot2::ggplot(data, ggplot2::aes(x, y)) + ggplot2::geom_text(ggplot2::aes(label = label))
ggplot2 		Label each point with a variable
Differs from Stata: The R example implements the graphical intent; size, jitter and theme scales are not numerically identical to Stata.
Stata reference · R reference 
scatter y x, msize(large)
 
ggplot2::ggplot(data, ggplot2::aes(x, y)) + ggplot2::geom_point(size = 3)
ggplot2 		Increase marker size
Differs from Stata: The R example implements the graphical intent; size, jitter and theme scales are not numerically identical to Stata.
Stata reference · R reference 
scatter y x, msymbol(Oh)
 
ggplot2::ggplot(data, ggplot2::aes(x, y)) + ggplot2::geom_point(shape = 1)
ggplot2 		Draw hollow circular markers
Differs from Stata: The R example implements the graphical intent; size, jitter and theme scales are not numerically identical to Stata.
Stata reference · R reference 
scatter y x, note("Source: survey")
 
ggplot2::ggplot(data, ggplot2::aes(x, y)) + ggplot2::geom_point() + ggplot2::labs(caption = "Source: survey")
ggplot2 		Add a source note to a graph
Differs from Stata: The R example implements the graphical intent; size, jitter and theme scales are not numerically identical to Stata.
Stata reference · R reference 
scatter y x, scheme(s2mono)
 
ggplot2::ggplot(data, ggplot2::aes(x, y)) + ggplot2::geom_point() + ggplot2::theme_bw()
ggplot2 		Use a monochrome-friendly theme
Differs from Stata: The R example implements the graphical intent; size, jitter and theme scales are not numerically identical to Stata.
Stata reference · R reference 
scatter y x, subtitle("Wave 1")
 
ggplot2::ggplot(data, ggplot2::aes(x, y)) + ggplot2::geom_point() + ggplot2::labs(subtitle = "Wave 1")
ggplot2 		Add a graph subtitle
Differs from Stata: The R example implements the graphical intent; size, jitter and theme scales are not numerically identical to Stata.
Stata reference · R reference 
scatter y x, title("Sample")
 
ggplot2::ggplot(data, ggplot2::aes(x, y)) + ggplot2::geom_point() + ggplot2::labs(title = "Sample")
ggplot2 		Add a graph title
Differs from Stata: The R example implements the graphical intent; size, jitter and theme scales are not numerically identical to Stata.
Stata reference · R reference 
scatter y x, xlabel(0(5)30)
 
ggplot2::ggplot(data, ggplot2::aes(x, y)) + ggplot2::geom_point() + ggplot2::scale_x_continuous(breaks = seq(0, 30, 5))
ggplot2 		Choose horizontal axis breaks
Differs from Stata: The R example implements the graphical intent; size, jitter and theme scales are not numerically identical to Stata.
Stata reference · R reference 
scatter y x, xline(10)
 
ggplot2::ggplot(data, ggplot2::aes(x, y)) + ggplot2::geom_point() + ggplot2::geom_vline(xintercept = 10)
ggplot2 		Add a vertical reference line
Differs from Stata: The R example implements the graphical intent; size, jitter and theme scales are not numerically identical to Stata.
Stata reference · R reference 
scatter y x, xscale(log)
 
ggplot2::ggplot(data, ggplot2::aes(x, y)) + ggplot2::geom_point() + ggplot2::scale_x_log10()
ggplot2 		Use a logarithmic predictor axis
Differs from Stata: The R example implements the graphical intent; size, jitter and theme scales are not numerically identical to Stata.
Stata reference · R reference 
scatter y x, xtitle("Age")
 
ggplot2::ggplot(data, ggplot2::aes(x, y)) + ggplot2::geom_point() + ggplot2::labs(x = "Age")
ggplot2 		Label the horizontal axis
Differs from Stata: The R example implements the graphical intent; size, jitter and theme scales are not numerically identical to Stata.
Stata reference · R reference 
scatter y x, ylabel(0(5)20)
 
ggplot2::ggplot(data, ggplot2::aes(x, y)) + ggplot2::geom_point() + ggplot2::scale_y_continuous(breaks = seq(0, 20, 5))
ggplot2 		Choose vertical axis breaks
Differs from Stata: The R example implements the graphical intent; size, jitter and theme scales are not numerically identical to Stata.
Stata reference · R reference 
scatter y x, yline(5)
 
ggplot2::ggplot(data, ggplot2::aes(x, y)) + ggplot2::geom_point() + ggplot2::geom_hline(yintercept = 5)
ggplot2 		Add a horizontal reference line
Differs from Stata: The R example implements the graphical intent; size, jitter and theme scales are not numerically identical to Stata.
Stata reference · R reference 
scatter y x, yscale(log)
 
ggplot2::ggplot(data, ggplot2::aes(x, y)) + ggplot2::geom_point() + ggplot2::scale_y_log10()
ggplot2 		Use a logarithmic response axis
Differs from Stata: The R example implements the graphical intent; size, jitter and theme scales are not numerically identical to Stata.
Stata reference · R reference 
scatter y x, ytitle("Income")
 
ggplot2::ggplot(data, ggplot2::aes(x, y)) + ggplot2::geom_point() + ggplot2::labs(y = "Income")
ggplot2 		Label the vertical axis
Differs from Stata: The R example implements the graphical intent; size, jitter and theme scales are not numerically identical to Stata.
Stata reference · R reference 
serrbar y se x
 
ggplot2::ggplot(data, ggplot2::aes(x, y)) + ggplot2::geom_errorbar(ggplot2::aes(ymin = y - se, ymax = y + se)) + ggplot2::geom_point()
ggplot2 		Plot supplied estimates and standard errors
Differs from Stata: Here se is a supplied standard-error variable; Stata serrbar defaults to one standard error. Rename the variable to suit your data.
Stata reference · R reference 
set scheme s1mono
 
theme_set(theme_bw())
ggplot2 		Change the look of every graph
Stata reference · R reference 
spikeplot z
 
{ counts <- as.data.frame(table(data$z)); graphics::plot(as.numeric(as.character(counts$Var1)), counts$Freq, type = "h") }
graphics 		Plot spikes for the frequencies of distinct values
Differs from Stata: Draws one spike for each observed value with its count. Axis ticks, spike widths, handling of missing values and graphics device styling differ from Stata spikeplot.
Stata reference · R reference 
symplot y
 
{ values <- sort(data$y); half <- length(values) %/% 2L; centre <- median(values); graphics::plot(centre - rev(head(values, half)), tail(values, half) - centre) }
graphics 		Check symmetry around the sample median
Differs from Stata: A diagnostic plotting workflow; plotting-position and standardization defaults can differ from Stata.
Stata reference · R reference 
twoway (scatter y x) (lfit y x)
 
ggplot(data, aes(x, y)) + geom_point() + geom_smooth(method = "lm")
ggplot2 		Scatter plot with a fitted line
Differs from Stata: geom_smooth(method = "lm") adds a confidence band by default; lfit has no band. Use se = FALSE for the same plotted layer.
Stata reference · R reference 
twoway area y x
 
ggplot(data, aes(x, y)) + geom_area()
ggplot2 		Area plot
Stata reference · R reference 
twoway connected y x
 
ggplot(data, aes(x, y)) + geom_line() + geom_point()
ggplot2 		Connected line plot
Differs from Stata: geom_line() sorts by x within groups; Stata line/connected uses observation order without the sort option.
Stata reference · R reference 
twoway rarea hi lo x
 
ggplot(data, aes(x, ymin = lo, ymax = hi)) + geom_ribbon()
ggplot2 		Shaded range
Stata reference · R reference 
twoway rcap hi lo x
 
ggplot(data, aes(x, ymin = lo, ymax = hi)) + geom_errorbar()
ggplot2 		Range plot with capped spikes
Stata reference · R reference 
twoway (scatter y x if g == 1) (scatter y x if g == 2)
 
ggplot(data, aes(x, y, colour = factor(g))) + geom_point()
ggplot2 		Colour points by group
Stata reference · R reference 
twoway lfitci y x
 
ggplot(data, aes(x, y)) + geom_smooth(method = "lm")
ggplot2 		Linear fit with a confidence band
Stata reference · R reference 
twoway qfit y x
 
ggplot(data, aes(x, y)) + geom_smooth(method = "lm", formula = y ~ poly(x, 2), se = FALSE)
ggplot2 		Quadratic fit
Stata reference · R reference 
twoway scatter y x, by(group)
 
ggplot(data, aes(x, y)) + geom_point() + facet_wrap(~group)
ggplot2 		One panel per group
Stata reference · R reference 
twoway (scatter y x) (lfit y x)
 
ggplot2::ggplot(data, ggplot2::aes(x, y)) + ggplot2::geom_point() + ggplot2::geom_smooth(method = "lm", formula = y ~ x, se = FALSE)
ggplot2 		Overlay a linear fitted relationship
Differs from Stata: Graphical workflow; plotting defaults can differ.
Stata reference · R reference 
twoway area y x, sort
 
ggplot2::ggplot(data, ggplot2::aes(x, y)) + ggplot2::geom_area()
ggplot2 		Fill the area beneath a line
Differs from Stata: Graphical workflow; plotting defaults can differ.
Stata reference · R reference 
twoway connected y x, sort
 
ggplot2::ggplot(data, ggplot2::aes(x, y)) + ggplot2::geom_line() + ggplot2::geom_point()
ggplot2 		Connect points in horizontal order
Differs from Stata: Graphical workflow; plotting defaults can differ.
Stata reference · R reference 
twoway contour y x z
 
ggplot2::ggplot(data, ggplot2::aes(x, y)) + ggplot2::geom_contour(ggplot2::aes(z = z))
ggplot2 		Draw contour lines on a regular grid
Differs from Stata: Graphical workflow; plotting defaults can differ.
Stata reference · R reference 
twoway dropline y x
 
ggplot2::ggplot(data, ggplot2::aes(x, y)) + ggplot2::geom_segment(ggplot2::aes(xend = x, yend = 0)) + ggplot2::geom_point()
ggplot2 		Draw points with drop lines
Differs from Stata: Graphical workflow; plotting defaults can differ.
Stata reference · R reference 
twoway lfitci y x
 
ggplot2::ggplot(data, ggplot2::aes(x, y)) + ggplot2::geom_smooth(method = "lm", formula = y ~ x, level = 0.95)
ggplot2 		Plot a linear fit with confidence limits
Differs from Stata: Graphical workflow; plotting defaults can differ.
Stata reference · R reference 
twoway pcarrow y x hi z
 
ggplot2::ggplot(data, ggplot2::aes(x, y)) + ggplot2::geom_segment(ggplot2::aes(xend = z, yend = hi), arrow = grid::arrow())
ggplot2 		Draw arrows between paired coordinates
Differs from Stata: Graphical workflow; plotting defaults can differ.
Stata reference · R reference 
twoway pcspike y x hi z
 
ggplot2::ggplot(data, ggplot2::aes(x, y)) + ggplot2::geom_segment(ggplot2::aes(xend = z, yend = hi))
ggplot2 		Connect paired coordinate endpoints
Differs from Stata: Graphical workflow; plotting defaults can differ.
Stata reference · R reference 
twoway qfitci y x
 
ggplot2::ggplot(data, ggplot2::aes(x, y)) + ggplot2::geom_smooth(method = "lm", formula = y ~ poly(x, 2), level = 0.95)
ggplot2 		Plot a quadratic fit with confidence limits
Differs from Stata: Graphical workflow; plotting defaults can differ.
Stata reference · R reference 
twoway rarea lo hi x, sort
 
ggplot2::ggplot(data, ggplot2::aes(x, y)) + ggplot2::geom_ribbon(ggplot2::aes(ymin = lo, ymax = hi))
ggplot2 		Draw a shaded uncertainty band
Differs from Stata: Graphical workflow; plotting defaults can differ.
Stata reference · R reference 
twoway rbar lo hi x
 
ggplot2::ggplot(data, ggplot2::aes(x, y)) + ggplot2::geom_rect(ggplot2::aes(xmin = x - 0.4, xmax = x + 0.4, ymin = lo, ymax = hi))
ggplot2 		Draw bars spanning lower and upper values
Differs from Stata: Graphical workflow; plotting defaults can differ.
Stata reference · R reference 
twoway rcap lo hi x
 
ggplot2::ggplot(data, ggplot2::aes(x, y)) + ggplot2::geom_errorbar(ggplot2::aes(ymin = lo, ymax = hi))
ggplot2 		Draw capped uncertainty intervals
Differs from Stata: Graphical workflow; plotting defaults can differ.
Stata reference · R reference 
twoway rscatter lo hi x
 
ggplot2::ggplot(data, ggplot2::aes(x, y)) + ggplot2::geom_point(ggplot2::aes(y = lo)) + ggplot2::geom_point(ggplot2::aes(y = hi))
ggplot2 		Plot both endpoints of a range
Differs from Stata: Graphical workflow; plotting defaults can differ.
Stata reference · R reference 
twoway rspike lo hi x
 
ggplot2::ggplot(data, ggplot2::aes(x, y)) + ggplot2::geom_linerange(ggplot2::aes(ymin = lo, ymax = hi))
ggplot2 		Show ranges with vertical lines
Differs from Stata: Graphical workflow; plotting defaults can differ.
Stata reference · R reference 
twoway scatter y x
 
ggplot2::ggplot(data, ggplot2::aes(x, y)) + ggplot2::geom_point()
ggplot2 		Draw a two-variable scatter plot
Differs from Stata: Graphical workflow; plotting defaults can differ.
Stata reference · R reference 
twoway spike y x
 
ggplot2::ggplot(data, ggplot2::aes(x, y)) + ggplot2::geom_segment(ggplot2::aes(xend = x, yend = 0))
ggplot2 		Draw vertical spikes from zero
Differs from Stata: Graphical workflow; plotting defaults can differ.
Stata reference · R reference 
vioplot y, over(group)
 
ggplot(data, aes(group, y)) + geom_violin()
ggplot2 		Violin plot
Stata reference · R reference

Programming utilities

Stata R Purpose, qualifications and sources 
about
 
utils::sessionInfo()
utils 		Inspect R version and platform information
Reports R session metadata; Stata license and serial information have no R counterpart.
Stata reference · R reference 
adopath
 
.libPaths()
base 		Inspect the package library search path
R package libraries replace ado search directories; this only reads the current path.
Stata reference · R reference 
program define addtwo
args a b
display `a'+`b'
end
 
addtwo <- function(a,b) a+b
base 		Receive two positional arguments in a function
R function arguments replace positional Stata macros; this is a fixed program example.
Stata reference · R reference 
forvalues i=1/4 {
    if `i'==2 continue
    display `i'
}
 
out <- integer(); for (i in 1:4) { if(i==2) next; out <- c(out,i) }
base 		Skip the rest of one loop iteration
Fixed workflow example; adapt the object and column names together.
Stata reference · R reference 
forvalues i=1/4 {
    if `i'==3 continue, break
    display `i'
}
 
out <- integer(); for (i in 1:4) { if(i==3) break; out <- c(out,i) }
base 		Exit a loop early
Fixed workflow example; adapt the object and column names together.
Stata reference · R reference 
creturn list
 
list(version=R.version.string,platform=R.version$platform,working_directory=getwd(),machine=.Machine)
base 		Inspect R version, platform, paths and numeric limits
Returns R's own settings and constants, not Stata's c-class values.
Stata reference · R reference 
dataex id x in 1/3
 
example <- data[1:3,c('id','x')]; text <- utils::capture.output(dput(example))
base 		Create reproducible R code for a small data example
dput emits R reconstruction code, not Stata input code or forum markup. Share only data already approved for disclosure.
Stata reference · R reference 
error 198
 
fail <- function() stop('Invalid input',call.=FALSE)
base 		Raise a clear error from a function
R conditions replace numeric Stata return codes; the check catches the deliberately raised error.
Stata reference · R reference 
program define positive
args x
if `x'<=0 exit
display `x'
end
 
positive <- function(x) { if(x<=0) return(invisible(NULL)); x }
base 		Return early from a function
Function-level return only; quitting the R process would terminate the user's session.
Stata reference · R reference 
ftools, version
 
list(version=as.character(utils::packageVersion('dplyr')),available=requireNamespace('dplyr',quietly=TRUE))
utils 		Inspect the R data-manipulation package version and availability
Reports the chosen R workflow dependency, dplyr; it does not inspect or compile a Stata Mata library.
Stata reference · R reference 
gettoken first rest : text
 
parts <- strsplit(text,' +')[[1]]; first <- parts[1]; rest <- paste(parts[-1],collapse=' ')
base 		Split the first simple token from the remaining text
Differs from Stata: Simple unquoted whitespace tokens only; Stata quoting and bind rules are not reproduced.
Stata reference · R reference 
include helper.do
 
sys.source(path,envir=environment())
base 		Execute helper code in the current evaluation environment
Only sources the controlled temporary R helper here; R syntax replaces the Stata do-file.
Stata reference · R reference 
mark touse if g == 2
 
data$touse <- as.integer(!is.na(data$g) & data$g==2)
base 		Create an inclusion marker from a condition
Stata reference · R reference 
markin if g == 2
 
selected <- which(data$g==2); interval <- range(selected)
base 		Find the row range enclosing a selected sample
Returns enclosing row endpoints only; selection can contain gaps and is not replaced by that interval.
Stata reference · R reference 
markout touse x y
 
data$touse <- as.integer(data$touse==1 & stats::complete.cases(data[c('x','y')]))
base 		Clear an inclusion marker when variables are incomplete
Stata reference · R reference 
syntax varlist
marksample touse
 
touse <- stats::complete.cases(data[vars])
base 		Mark complete rows over supplied column names
R receives column names explicitly; no Stata syntax parser state or if/in qualifiers are read.
Stata reference · R reference 
matlist X, format(%9.2f)
 
format(X,nsmall=2,trim=TRUE)
base 		Format a numeric matrix for display
Differs from Stata: Returns a character matrix; R does not duplicate Stata's borders/equation headers.
Stata reference · R reference 
matname X first second, columns(1..2)
 
colnames(X) <- c('first','second')
base 		Name matrix columns
Fixed workflow example; adapt the object and column names together.
Stata reference · R reference 
mkmat x y, matrix(X)
 
X <- as.matrix(data[c('x','y')])
base 		Create a matrix from selected numeric variables
Differs from Stata: X is the fixed matrix object name; R row names need not match Stata r1, r2 labels.
Stata reference · R reference 
mkmat x y, matrix(X) nomissing
 
X <- as.matrix(data[stats::complete.cases(data[c('x','y')]),c('x','y')])
base 		Create a matrix after listwise missing exclusion
Stata reference · R reference 
numlist "1(2)9"
 
values <- seq(1,9,by=2)
base 		Expand a numeric list with a fixed step
Constructs this list directly; it does not parse arbitrary Stata numlist strings.
Stata reference · R reference 
numlist "5/1"
 
values <- seq(5,1,by=-1)
base 		Expand a decreasing consecutive numeric list
Direct R sequence workflow for the displayed Stata list.
Stata reference · R reference 
post handle (1) (2.5)
 
results <- rbind(results,data.frame(id=1L,value=2.5))
base 		Append one result observation
Appends to the R accumulator; a separate save writes the final file.
Stata reference · R reference 
postclose handle
 
saveRDS(results,path)
base 		Persist an accumulated result table
Finalizes the R accumulator as RDS rather than closing a Stata postfile handle.
Stata reference · R reference 
postfile handle id value using results.dta, replace
 
results <- data.frame(id=integer(),value=double())
base 		Initialize a typed result accumulator
R accumulator workflow: append rows, then save separately; no disk handle is opened yet.
Stata reference · R reference 
query output
 
options()[c('width','digits','scipen')]
base 		Inspect current R output settings
Read-only R settings query; these are R-specific options rather than Stata parameter names.
Stata reference · R reference 
require ftools
 
stopifnot(requireNamespace('dplyr',quietly=TRUE))
base 		Require an installed R workflow dependency
Checks the R counterpart dependency dplyr; no package is installed and Stata package names do not translate automatically.
Stata reference · R reference 
require ftools >= 2.48.0
 
stopifnot(requireNamespace('dplyr',quietly=TRUE),utils::packageVersion('dplyr') >= '1.1.0')
base 		Require a minimum version of an R workflow dependency
The R counterpart has its own version requirement; 1.1.0 supports the .by and relationship features used in these examples.
Stata reference · R reference 
sleep 1
 
Sys.sleep(0.001)
base 		Pause execution using milliseconds converted to seconds
R Sys.sleep takes seconds; no machine-speed timing assertion is made.
Stata reference · R reference 
svmat X, names(col)
 
data <- as.data.frame(X)
base 		Convert matrix columns to named variables
Differs from Stata: Creates a new table; merging with an existing dataset requires an explicit row/key alignment.
Stata reference · R reference 
syntax varname(numeric) [, Scale(real 1)]
 
scaled_sum <- function(data,var,scale=1) { stopifnot(length(var)==1L,var %in% names(data),is.numeric(data[[var]]),is.numeric(scale),length(scale)==1L); sum(data[[var]],na.rm=TRUE)*scale }
base 		Validate a numeric column argument and an optional scalar
Uses R formals and explicit checks; it does not parse Stata's command grammar.
Stata reference · R reference 
sysdir list
 
list(home=R.home(),library=R.home('library'),temporary=tempdir())
base 		Inspect R installation and temporary directories
Shows R directories; Stata BASE/PLUS/PERSONAL directories have no one-to-one mapping.
Stata reference · R reference 
tempfile cache
 
path <- tempfile(fileext='.rds')
base 		Reserve a unique temporary path
Differs from Stata: tempfile() reserves a name without creating a file; R does not guarantee Stata-style cleanup of each named file when a function exits.
Stata reference · R reference 
tempname total
scalar `total' = 10
 
local({ total <- 10; total*2 })
base 		Use a temporary scalar inside local scope
Lexical scope supplies isolation rather than a unique Stata temporary object name.
Stata reference · R reference 
tempvar squared
generate `squared' = x^2
 
local({ squared <- data$x^2; sum(squared,na.rm=TRUE) })
base 		Compute with a temporary vector in a local scope
R local scope avoids adding a temporary dataset column; no Stata macro name is generated.
Stata reference · R reference 
tokenize "alpha beta" gamma
 
tokens <- scan(text='"alpha beta" gamma',what=character(),quiet=TRUE)
base 		Tokenize a quoted phrase and a word
Differs from Stata: Demonstrates basic double-quoted whitespace tokens; Stata compound quoting has no automatic equivalent.
Stata reference · R reference 
tokenize alpha beta gamma
 
tokens <- strsplit('alpha beta gamma',' +')[[1]]
base 		Split simple whitespace-separated tokens
Differs from Stata: R returns a character vector instead of positional local macros; quoted-token parsing differs.
Stata reference · R reference 
unab vars : income
 
vars <- names(data)[pmatch('income',names(data),duplicates.ok=FALSE)]; stopifnot(!anyNA(vars))
base 		Resolve an unambiguous abbreviated column name
R partial matching must be requested explicitly; exact matches take precedence.
Stata reference · R reference 
unab vars : x*
 
vars <- grep('^x',names(data),value=TRUE)
base 		Expand a column-name prefix wildcard
Handles the displayed simple prefix wildcard; this is not a complete Stata varlist parser.
Stata reference · R reference 
unabcmd gen
 
full <- match.arg('mea',c('mean','median','sum'))
base 		Resolve a unique abbreviation among known R choices
Illustrates explicit R-choice matching, not translation of Stata command abbreviations; Stata gen corresponds to generate.
Stata reference · R reference 
viewsource mean.ado
 
body(base::mean.default)
base 		Inspect the body of a named R function
Reads an R function body rather than opening a Stata ado source editor.
Stata reference · R reference

Data management

Stata R Purpose, qualifications and sources 
append using other.dta
 
data <- dplyr::bind_rows(data,other)
dplyr 		Append tables allowing different column sets
Fixed workflow example; adapt the object and column names together.
Stata reference · R reference 
append using other.dta, keep(x)
 
data <- dplyr::bind_rows(data,other['x'])
dplyr 		Append only selected columns from the using table
Stata reference · R reference 
append using second.dta third.dta, generate(source)
 
data <- dplyr::bind_rows(list('0'=data,'1'=second,'2'=third),.id='source'); data$source <- as.integer(data$source)
dplyr 		Record origin while appending several datasets
Stata reference · R reference 
assertnested id, within(g)
 
stopifnot(all(vapply(split(data$g,data$id),function(z) length(unique(z))==1L,logical(1))))
base 		Assert a group attribute is constant for each identifier
Stata reference · R reference 
assertnested region district
 
stopifnot(all(vapply(split(data$region,data$district),function(z) length(unique(z))==1L,logical(1))))
base 		Assert that districts belong to one region
Stata reference · R reference 
cf x using other.dta, verbose
 
which(data$x != other$x | xor(is.na(data$x),is.na(other$x)))
base 		Locate differing rows in a selected column
Workflow inspects x only; compare each column for the complete verbose report.
Stata reference · R reference 
cf x y using other.dta
 
stopifnot(identical(unname(as.matrix(data[c('x','y')])),unname(as.matrix(other[c('x','y')]))))
base 		Assert selected dataset values match
Differs from Stata: Compares the supplied in-memory other table in row order; ignores metadata as cf does.
Stata reference · R reference 
clonevar copy = x
 
data$copy <- data$x
base 		Copy a column and its attributes
Stata reference · R reference 
codebook x, compact
 
c(nonmissing=sum(!is.na(data$x)),unique=length(unique(data$x[!is.na(data$x)])),min=min(data$x,na.rm=TRUE),max=max(data$x,na.rm=TRUE))
base 		Describe observed count, distinct count and range
Differs from Stata: Numeric codebook subset; does not reproduce Stata's variable-label and format report.
Stata reference · R reference 
compare x y
 
c(less=sum(data$x<data$y,na.rm=TRUE),equal=sum(data$x==data$y,na.rm=TRUE),greater=sum(data$x>data$y,na.rm=TRUE),x_missing_only=sum(is.na(data$x)&!is.na(data$y)),both_missing=sum(is.na(data$x)&is.na(data$y)))
base 		Compare value ordering and missingness
Differs from Stata: Produces the principal comparison counts; not Stata's formatted difference summary.
Stata reference · R reference 
compare x y if g == 1
 
with(subset(data,g==1),table(sign(x-y),useNA='ifany'))
base 		Compare ordering inside a subset
Differs from Stata: Returns ordering counts only; use a separate summary for difference magnitudes.
Stata reference · R reference 
compress x
 
stopifnot(all(is.na(data$x) | (data$x==trunc(data$x) & abs(data$x)<=.Machine$integer.max))); data$x <- as.integer(data$x)
base 		Reduce safe whole-number storage to R integers
Manual whole-number optimization only. R has no Stata byte/int storage classes or general compress equivalent.
Stata reference · R reference 
count if !missing(x, y)
 
sum(stats::complete.cases(data[c('x','y')]))
base 		Count rows complete on selected columns
Stata reference · R reference 
count if missing(x)
 
sum(is.na(data$x))
base 		Count missing values
Stata reference · R reference 
count if x > 2 & x < .
 
sum(data$x > 2,na.rm=TRUE)
base 		Count observed values above a threshold
Stata reference · R reference 
count in 2/5
 
nrow(data[2:5,,drop=FALSE])
base 		Count a row interval
Fixed workflow example; adapt the object and column names together.
Stata reference · R reference 
cross using other.dta
 
data <- dplyr::cross_join(data,other)
dplyr 		Form the Cartesian product of two datasets
Fixed workflow example; adapt the object and column names together.
Stata reference · R reference 
cumul x, generate(c)
 
data$c <- dplyr::row_number(data$x) / sum(!is.na(data$x))
dplyr 		Create the empirical cumulative distribution of a variable
Differs from Stata: Missing values stay missing and are excluded from the denominator. By default Stata gives tied values distinct cumulatives in an arbitrary order; R breaks ties by row order. Weights and by() need an explicit grouped or weighted version.
Stata reference · R reference 
cumul x, generate(c) equal
 
data$c <- dplyr::cume_dist(data$x)
dplyr 		Cumulative distribution with equal values for ties
Tied values share the largest cumulative proportion, as ecdf(x)(x) does; missing values stay missing.
Stata reference · R reference 
datasignature confirm
 
stopifnot(identical(signature,serialize(data,NULL,version=2)))
base 		Assert data match a stored R signature
R-specific byte comparison; not interoperable with Stata .dtasig files.
Stata reference · R reference 
datasignature set
 
signature <- serialize(data,NULL,version=2)
base 		Store an exact R serialization for later comparison
R-specific serialized snapshot; not Stata's checksum algorithm, and R metadata is included.
Stata reference · R reference 
describe x y
 
vapply(data[c('x','y')],typeof,character(1))
base 		Inspect storage types of selected columns
Stata reference · R reference 
describe, short
 
c(observations=nrow(data),variables=ncol(data))
base 		Get dataset dimensions compactly
Fixed workflow example; adapt the object and column names together.
Stata reference · R reference 
destring s, generate(x)
 
data$x <- as.numeric(data$s)
base 		Parse clean numeric strings
Stata reference · R reference 
destring s, generate(x) dpcomma
 
data$x <- as.numeric(sub(',','.',data$s,fixed=TRUE))
base 		Parse a decimal-comma field
Stata reference · R reference 
destring s, generate(x) ignore(",")
 
data$x <- as.numeric(gsub(',','',data$s,fixed=TRUE))
base 		Parse numbers containing grouping commas
Stata reference · R reference 
destring s, generate(x) percent
 
data$x <- as.numeric(sub('%$','',data$s))/100
base 		Convert percentage text to proportions
Stata reference · R reference 
drop if g == 2
 
data <- data[data$g != 2, , drop=FALSE]
base 		Drop a category of rows
Stata reference · R reference 
drop if missing(x)
 
data <- data[!is.na(data$x), , drop=FALSE]
base 		Drop rows missing a value
Stata reference · R reference 
drop if missing(x, y)
 
data <- tidyr::drop_na(data, x, y)
tidyr 		Drop rows incomplete on selected columns
Stata reference · R reference 
drop if x <= 2
 
data <- dplyr::filter(data, is.na(x) | x > 2)
dplyr 		Drop low values while retaining missing rows
Differs from Stata: Stata numeric missing sorts above finite values; the explicit is.na branch preserves those rows.
Stata reference · R reference 
drop in 2/4
 
data <- data[-(2:4), , drop=FALSE]
base 		Drop a row interval
Fixed workflow example; adapt the object and column names together.
Stata reference · R reference 
drop s y
 
data <- data[setdiff(names(data),c('s','y'))]
base 		Remove selected columns
Stata reference · R reference 
ds id g, not
 
setdiff(names(data),c('id','g'))
base 		Find columns outside an exclusion list
Stata reference · R reference 
ds, has(type numeric)
 
names(data)[vapply(data,is.numeric,logical(1))]
base 		Find names of numeric columns
Fixed workflow example; adapt the object and column names together.
Stata reference · R reference 
ds, has(type string)
 
names(data)[vapply(data,is.character,logical(1))]
base 		Find names of character columns
Fixed workflow example; adapt the object and column names together.
Stata reference · R reference 
duplicates drop id, force
 
data <- data[!duplicated(data$id),,drop=FALSE]
base 		Retain the first observation per duplicated key
Stata reference · R reference 
duplicates list id
 
data[duplicated(data$id)|duplicated(data$id,fromLast=TRUE),,drop=FALSE]
base 		Inspect all members of duplicated keys
Stata reference · R reference 
duplicates report id
 
counts <- as.integer(table(data$id)); data.frame(copies=sort(unique(counts)),observations=vapply(sort(unique(counts)),function(k) sum(counts[counts==k]),integer(1)),surplus=vapply(sort(unique(counts)),function(k) sum(counts==k)*(k-1L),integer(1)))
base 		Summarize copies, observations and duplicate surplus
Stata reference · R reference 
duplicates tag id, generate(dup)
 
data <- dplyr::mutate(data,dup=dplyr::n()-1L,.by=id)
dplyr 		Tag all group members with duplicate surplus count
Stata reference · R reference 
bysort g: egen z = mean(x)
 
data <- dplyr::mutate(data,z=mean(x,na.rm=TRUE),.by=g)
dplyr 		Attach group means without collapsing rows
Stata reference · R reference 
bysort g: egen z = total(x)
 
data <- dplyr::mutate(data,z=sum(x,na.rm=TRUE),.by=g)
dplyr 		Attach group totals without collapsing rows
Stata reference · R reference 
egen z = anycount(x y), values(2 4)
 
data$z <- rowSums(cbind(data$x %in% c(2,4),data$y %in% c(2,4)))
base 		Count listed values across columns
Stata reference · R reference 
egen z = anymatch(x y), values(2 4)
 
data$z <- as.integer(data$x %in% c(2,4) | data$y %in% c(2,4))
base 		Flag any listed value across columns
Stata reference · R reference 
egen z = concat(g s), punct("-")
 
data$z <- paste(data$g,ifelse(is.na(data$s),'',data$s),sep='-')
base 		Concatenate numeric and string columns
Differs from Stata: Treats R NA strings as Stata empty strings; numeric formatting is simple decimal here.
Stata reference · R reference 
egen z = count(x)
 
data$z <- sum(!is.na(data$x))
base 		Repeat the nonmissing count in a column
Stata reference · R reference 
egen z = cut(x), at(0,3,7)
 
data$z <- as.numeric(as.character(cut(data$x,breaks=c(0,3,7),right=FALSE,labels=c(0,3))))
base 		Bin values and label bins by lower endpoint
Stata reference · R reference 
egen z = diff(x y)
 
data$z <- as.integer(!(data$x==data$y | (is.na(data$x)&is.na(data$y)))); data$z[xor(is.na(data$x),is.na(data$y))] <- 1L
base 		Flag differing values across two columns
Differs from Stata: Ordinary missing is handled explicitly; extended Stata missing categories need separate comparison.
Stata reference · R reference 
egen z = firstnm(x)
 
data$z <- data$x[which(!is.na(data$x))[1]]
base 		Repeat the first nonmissing value
Stata reference · R reference 
egen z = group(g s)
 
data$z <- as.integer(interaction(data$g,data$s,drop=TRUE,lex.order=TRUE)); data$z[is.na(data$s) | data$s==''] <- NA_integer_
base 		Encode combinations of group keys
Differs from Stata: Combination IDs identify groups; exact numeric codes and string collation may differ from Stata.
Stata reference · R reference 
egen z = iqr(x)
 
data$z <- diff(stats::quantile(data$x,c(.25,.75),na.rm=TRUE,type=2))
base 		Repeat the interquartile range
Differs from Stata: Uses R quantile type 2 for Stata's default empirical percentile rule.
Stata reference · R reference 
egen z = lastnm(x)
 
data$z <- utils::tail(c(NA_real_, data$x[!is.na(data$x)]), 1)
base 		Repeat the last nonmissing value
Stata reference · R reference 
egen z = max(x)
 
data$z <- if (all(is.na(data$x))) NA_real_ else max(data$x, na.rm=TRUE)
base 		Repeat the maximum in a column
Stata reference · R reference 
egen z = mean(x)
 
data$z <- mean(data$x,na.rm=TRUE)
base 		Repeat the mean in a column
Stata reference · R reference 
egen z = median(x)
 
data$z <- stats::median(data$x,na.rm=TRUE)
base 		Repeat the median in a column
Stata reference · R reference 
egen z = min(x)
 
data$z <- if (all(is.na(data$x))) NA_real_ else min(data$x, na.rm=TRUE)
base 		Repeat the minimum in a column
Stata reference · R reference 
egen z = pc(x)
 
data$z <- 100*data$x/sum(data$x,na.rm=TRUE)
base 		Express values as percentages of the column total
Requires a nonzero observed total; this is a share calculation, not survey estimation.
Stata reference · R reference 
egen z = pctile(x), p(25)
 
data$z <- as.numeric(stats::quantile(data$x,.25,na.rm=TRUE,type=2))
base 		Repeat a specified empirical percentile
Differs from Stata: Uses R quantile type 2 for Stata's default empirical percentile rule.
Stata reference · R reference 
egen z = rank(x)
 
data$z <- rank(data$x,na.last='keep',ties.method='average')
base 		Rank observed values using average ties
Stata reference · R reference 
egen z = rank(x), field
 
data$z <- rank(-data$x,na.last='keep',ties.method='min')
base 		Rank highest value first with competition ties
Stata reference · R reference 
egen z = rowmax(x y)
 
data$z <- pmax(data$x,data$y,na.rm=TRUE)
base 		Find the available row maximum
Stata reference · R reference 
egen z = rowmean(x y)
 
data$z <- rowMeans(data[c('x','y')],na.rm=TRUE); data$z[is.nan(data$z)] <- NA_real_
base 		Average available values across columns
Stata reference · R reference 
egen z = rowmin(x y)
 
data$z <- pmin(data$x,data$y,na.rm=TRUE)
base 		Find the available row minimum
Stata reference · R reference 
egen z = rowmiss(x y)
 
data$z <- rowSums(is.na(data[c('x','y')]))
base 		Count missing cells in each row
Stata reference · R reference 
egen z = rownonmiss(x y)
 
data$z <- rowSums(!is.na(data[c('x','y')]))
base 		Count observed cells in each row
Stata reference · R reference 
egen z = rowtotal(x y)
 
data$z <- rowSums(data[c('x','y')],na.rm=TRUE)
base 		Sum available values across columns
Stata reference · R reference 
egen z = rowtotal(x y), missing
 
data$z <- rowSums(data[c('x','y')],na.rm=TRUE); data$z[rowSums(!is.na(data[c('x','y')]))==0] <- NA_real_
base 		Return missing for entirely missing row totals
Stata reference · R reference 
egen z = sd(x)
 
data$z <- stats::sd(data$x,na.rm=TRUE)
base 		Repeat the sample standard deviation in a column
Stata reference · R reference 
egen z = std(x)
 
data$z <- as.numeric(scale(data$x))
base 		Standardize using the sample standard deviation
Stata reference · R reference 
egen z = tag(g)
 
data$z <- as.integer(!is.na(data$g) & !duplicated(data$g))
base 		Tag the first observed row of each group
Stata reference · R reference 
egen z = total(x)
 
data$z <- sum(data$x,na.rm=TRUE)
base 		Repeat the sum in a column
Stata reference · R reference 
expandcl 2, cluster(g) generate(newg)
 
parts <- split(data,data$g); data <- do.call(rbind,lapply(seq_along(parts),function(k) do.call(rbind,lapply(1:2,function(j) transform(parts[[k]],newg=(k-1L)*2L+j)))))
base 		Duplicate whole groups with fresh group identifiers
Differs from Stata: Validates cluster membership and copy counts; the numeric assignment/order of new IDs may differ from Stata.
Stata reference · R reference 
fcollapse (mean) z=x, by(g)
 
data <- dplyr::summarise(data,z=mean(x,na.rm=TRUE),.by=g)
dplyr 		Collapse means with the ftools public helper
Stata reference · R reference 
fcollapse (mean) z=x, by(g) merge
 
data <- dplyr::mutate(data,z=mean(x,na.rm=TRUE),.by=g)
dplyr 		Merge fcollapse group summaries back into original rows
Stata reference · R reference 
fcollapse (nansum) z=x, by(g)
 
data <- dplyr::summarise(data,z=if(all(is.na(x))) NA_real_ else sum(x,na.rm=TRUE),.by=g)
dplyr 		Preserve missing for entirely missing group sums
Stata reference · R reference 
fillin g h
 
data <- tidyr::complete(data,g,h)
tidyr 		Insert absent combinations of categorical keys
Differs from Stata: Returns the rectangularized table without Stata's _fillin indicator; factor levels may add unused R categories.
Stata reference · R reference 
fillin g h
 
data$`_fillin` <- 0L; data <- tidyr::complete(data,g,h,fill=list(`_fillin`=1L))
tidyr 		Mark rows inserted during categorical rectangularization
Stata reference · R reference 
flevelsof g, local(groups)
 
groups <- sort(unique(data$g[!is.na(data$g)]))
base 		Store observed categories from the ftools levels helper
Stata reference · R reference 
fmerge 1:1 id using other.dta, nogenerate
 
data <- dplyr::full_join(data,other,by='id',relationship='one-to-one')
dplyr 		Perform a validated ftools-style one-to-one merge
Requires disjoint non-key column names unless update is explicitly shown. dplyr suffixes overlapping names; ordinary Stata merge preserves master values, including master missing values, on matched rows.
Stata reference · R reference 
fralias add score, from(link)
 
get_score <- function() lookup$score[match(data$id,lookup$id)]
base 		Read linked values dynamically without storing a copied column
R accessor-function workflow, not an alias column: call get_score() each time; modifications belong in lookup.
Stata reference · R reference 
frget newscore=score, from(link)
 
data$newscore <- lookup$score[data$link]
base 		Rename a linked value when copying it
Value-copy workflow; frame linkage and Stata metadata are not carried into R.
Stata reference · R reference 
frget score, from(link)
 
data$score <- lookup$score[data$link]
base 		Copy a value through an established table link
Differs from Stata: Copies values using the precomputed link. Stata display formats and metadata require explicit copying in R.
Stata reference · R reference 
frlink 1:1 id, frame(lookup)
 
stopifnot(!anyDuplicated(data$id),!anyDuplicated(lookup$id)); link <- match(data$id,lookup$id)
base 		Validate and construct a one-to-one table link
link is a fixed R object name; unmatched keys produce NA. Rebuild after source changes.
Stata reference · R reference 
frlink m:1 id, frame(lookup) generate(link)
 
stopifnot(!anyDuplicated(lookup$id)); data$link <- match(data$id,lookup$id)
base 		Build a many-to-one row link to a lookup table
R match indices are a snapshot and must be rebuilt after lookup row reordering.
Stata reference · R reference 
frunalias score
 
data$score <- lookup$score[match(data$id, lookup$id)]
base 		Materialize a dynamically looked-up value as a column
Materialize an ordinary R column snapshot. Requires data and lookup tables with matching id keys and a lookup score column; later lookup changes do not change this column.
Stata reference · R reference 
gcollapse (mean) z=x, by(g)
 
data <- dplyr::summarise(data,z=mean(x,na.rm=TRUE),.by=g)
dplyr 		Collapse group means using the gcollapse workflow
Stata reference · R reference 
gcollapse (mean) z=x, by(g) freq(n)
 
data <- dplyr::summarise(data,z=mean(x,na.rm=TRUE),n=dplyr::n(),.by=g)
dplyr 		Keep raw group row counts beside collapsed means
Stata reference · R reference 
gcollapse (mean) z=x, by(g) merge
 
data <- dplyr::mutate(data,z=mean(x,na.rm=TRUE),.by=g)
dplyr 		Add collapsed group means back to all original rows
Stata reference · R reference 
gcollapse (nmissing) n=x, by(g)
 
data <- dplyr::summarise(data,n=sum(is.na(x)),.by=g)
dplyr 		Collapse group missing-value counts
Stata reference · R reference 
gcollapse (nunique) n=x, by(g)
 
data <- dplyr::summarise(data,n=dplyr::n_distinct(x,na.rm=TRUE),.by=g)
dplyr 		Collapse group distinct-value counts on complete data
Differs from Stata: Complete values in this fixture; treatment of missing categories is not asserted.
Stata reference · R reference 
gcontract g h, freq(n)
 
data <- dplyr::count(data,g,h,name='n')
dplyr 		Count jointly observed category combinations
Stata reference · R reference 
gcontract g, freq(n) percent(p)
 
data <- dplyr::count(data,g,name='n'); data$p <- 100*data$n/sum(data$n)
dplyr 		Create category counts and percentages with gcontract
Stata reference · R reference 
gdistinct g h
 
vapply(data[c('g','h')],function(z) length(unique(z[!is.na(z)])),integer(1))
base 		Count distinct values separately for several columns
Stata reference · R reference 
gdistinct g h, joint
 
nrow(unique(data[stats::complete.cases(data[c('g','h')]),c('g','h')]))
base 		Count distinct complete combinations jointly
Stata reference · R reference 
gduplicates drop g h, force
 
data <- data[!duplicated(data[c('g','h')]),,drop=FALSE]
base 		Keep the first row of each compound key
Stata reference · R reference 
gduplicates tag g h, generate(dup)
 
data <- dplyr::mutate(data,dup=dplyr::n()-1L,.by=c(g,h))
dplyr 		Tag all duplicates of compound keys
Stata reference · R reference 
gegen groupid = group(g h)
 
data$groupid <- as.integer(interaction(data$g,data$h,drop=TRUE,lex.order=TRUE))
base 		Create identifiers for complete combinations of keys
Differs from Stata: Complete numeric keys only; exact identifier ordering follows the demonstrated lexicographic keys.
Stata reference · R reference 
gegen tag = tag(g)
 
data$tag <- as.integer(!is.na(data$g)&!duplicated(data$g))
base 		Tag one observation in each nonmissing group
Stata reference · R reference 
gegen z = mean(x), by(g)
 
data <- dplyr::mutate(data,z=mean(x,na.rm=TRUE),.by=g)
dplyr 		Attach group means using a gegen-style workflow
Stata reference · R reference 
bysort g (id): generate row = _n
 
data <- dplyr::arrange(data,g,id); data <- dplyr::mutate(data,row=dplyr::row_number(),.by=g)
dplyr 		Number observations within sorted groups
Stata reference · R reference 
bysort g (id): generate total = sum(x)
 
data <- dplyr::arrange(data,g,id); data <- dplyr::mutate(data,total=cumsum(replace(x,is.na(x),0)),.by=g)
dplyr 		Compute running totals within groups
Stata reference · R reference 
bysort g: generate n = _N
 
data <- dplyr::mutate(data,n=dplyr::n(),.by=g)
dplyr 		Attach group row counts
Stata reference · R reference 
bysort id (year): generate first = _n == 1
 
{ data <- dplyr::arrange(data, id, year); dplyr::mutate(data, first = as.integer(dplyr::row_number() == 1L), .by = id) }
dplyr 		Mark the first observation within sorted groups
Differs from Stata: Sorts by id and year, then marks one first row per id, including a missing-id group. Ties in year use input order in R; add a tie-break variable for a reproducible cross-system choice.
Stata reference · R reference 
generate byte high = x > 3 if x < .
 
data$high <- ifelse(is.na(data$x),NA_integer_,as.integer(data$x > 3))
base 		Create a binary indicator retaining missing values
Stata reference · R reference 
generate byte observed = !missing(x)
 
data$observed <- as.integer(!is.na(data$x))
base 		Create a nonmissing indicator
Stata reference · R reference 
generate cumulative = sum(x)
 
data$cumulative <- cumsum(replace(data$x,is.na(data$x),0))
base 		Create a running sum ignoring missing values
Stata's generate sum() treats missing contributions as zero, unlike raw cumsum().
Stata reference · R reference 
generate double z = x^2
 
data$z <- data$x^2
base 		Square a variable preserving missing values
Stata reference · R reference 
generate n = _N
 
data$n <- nrow(data)
base 		Repeat the total row count
Stata reference · R reference 
generate prior = x[_n-1]
 
data$prior <- dplyr::lag(data$x)
dplyr 		Shift a column by row position
Differs from Stata: Position-based shift; this does not use time-series spacing or tsset semantics.
Stata reference · R reference 
generate row = _n
 
data$row <- seq_len(nrow(data))
base 		Create row numbers
Stata reference · R reference 
generate z = x + y
 
data$z <- data$x + data$y
base 		Add columns with missing propagation
Stata reference · R reference 
getmata (x y)=X
 
data <- as.data.frame(X); names(data) <- c('x','y')
base 		Convert a matrix into named data columns
Creates a new R table rather than importing from a separate Mata workspace; X is fixed.
Stata reference · R reference 
gisid g h, missok
 
stopifnot(!anyDuplicated(data[c('g','h')]))
base 		Validate compound-key uniqueness allowing ordinary missing
Differs from Stata: Ordinary missing is allowed as a key value; extended Stata missing categories are not represented here.
Stata reference · R reference 
gisid id
 
stopifnot(!anyNA(data$id),!anyDuplicated(data$id))
base 		Assert a unique complete identifier with gisid
Stata reference · R reference 
glevelsof g h
 
levels <- unique(data[c('g','h')]); levels <- levels[do.call(order,levels),,drop=FALSE]
base 		List distinct complete multi-column keys
Differs from Stata: Returns a key table rather than a space-delimited local macro; fixture keys contain no missing values.
Stata reference · R reference 
glevelsof g, local(levels)
 
levels <- sort(unique(data$g[!is.na(data$g)]))
base 		Store sorted distinct categories with glevelsof
Stata reference · R reference 
gquantiles quartile=x, xtile nquantiles(4)
 
cuts <- stats::quantile(data$x,c(.25,.5,.75),type=2); data$quartile <- as.integer(cut(data$x,breaks=c(-Inf,cuts,Inf),right=TRUE))
base 		Assign four empirical quantile categories
Differs from Stata: Unweighted distinct finite values; uses explicit R type-2 cutoffs and right-closed categories. Ties can prevent equal group sizes.
Stata reference · R reference 
gquantiles x, _pctile percentiles(25 50 75)
 
stats::quantile(data$x,c(.25,.5,.75),type=2)
base 		Return selected empirical percentile cutpoints
Differs from Stata: Explicit R type-2 unweighted percentile workflow; no alternative-definition or weighted numerical parity is asserted.
Stata reference · R reference 
greshape gather x1 x2, keys(measure) values(value)
 
data <- tidyr::pivot_longer(data,c(x1,x2),names_to='measure',values_to='value')
tidyr 		Gather explicitly named wide columns with greshape
Stata reference · R reference 
greshape long x, by(id) keys(wave) dropmiss
 
data <- tidyr::pivot_longer(data,c(x1,x2),names_to='wave',names_prefix='x',names_transform=list(wave=as.integer),values_to='x',values_drop_na=TRUE)
tidyr 		Drop missing measurements while reshaping one stub
Differs from Stata: Single-stub fixed workflow; for multiple stubs drop a row only when every reshaped value is missing.
Stata reference · R reference 
greshape spread value, keys(measure)
 
data <- tidyr::pivot_wider(data,names_from=measure,values_from=value)
tidyr 		Spread text-keyed values into columns with greshape
Stata reference · R reference 
gsort -x
 
data <- data[order(-data$x,na.last=FALSE), , drop=FALSE]
base 		Sort descending with Stata missing first
Numeric missing sorts first in Stata descending order; use na.last=FALSE deliberately.
Stata reference · R reference 
gsort g -y
 
data <- dplyr::arrange(data, g, dplyr::desc(y))
dplyr 		Sort by ascending group and descending value
Stata reference · R reference 
gstats winsor x, by(g) cuts(25 75) replace
 
data <- dplyr::mutate(data,x={b <- stats::quantile(x,c(.25,.75),na.rm=TRUE,type=2); pmax(b[1],pmin(b[2],x))},.by=g)
dplyr 		Cap tails separately inside each group
Differs from Stata: Groups use independent unweighted R type-2 cutoffs; all-missing groups require an explicit policy.
Stata reference · R reference 
gstats winsor x, cuts(25 75) generate(z)
 
bounds <- stats::quantile(data$x,c(.25,.75),na.rm=TRUE,type=2); data$z <- pmax(bounds[1],pmin(bounds[2],data$x))
base 		Cap both tails at explicitly selected empirical quartiles
Differs from Stata: Unweighted workflow with an explicit R type-2 empirical percentile rule; no cross-software numerical certification is claimed.
Stata reference · R reference 
gstats winsor x, cuts(25 75) trim generate(z)
 
bounds <- stats::quantile(data$x,c(.25,.75),na.rm=TRUE,type=2); data$z <- data$x; data$z[!is.na(data$x) & (data$x<bounds[1] | data$x>bounds[2])] <- NA_real_
base 		Set values outside empirical quartiles to missing
Differs from Stata: Trimming sets tails to missing while retaining rows; percentile rule is explicit R type 2 and no weighted parity is asserted.
Stata reference · R reference 
gtoplevelsof g, ntop(2)
 
counts <- dplyr::count(data,g,name='n'); counts <- counts[order(-counts$n,counts$g),]; top <- utils::head(counts,2)
dplyr 		Find the two most frequent complete categories
Differs from Stata: Returns the leading categories and counts only; Stata cumulative/Other rows are not generated. gtop is an alias and is not counted separately.
Stata reference · R reference 
gunique g h
 
nrow(unique(data[c('g','h')]))
base 		Count unique complete row-key combinations
Differs from Stata: Complete keys in this fixture; unlike gdistinct without joint, gunique treats all listed columns jointly.
Stata reference · R reference 
gunique h, by(g) generate(nunique)
 
data <- dplyr::mutate(data,nunique=dplyr::n_distinct(h),.by=g)
dplyr 		Attach distinct-value counts within each group
Differs from Stata: Complete group and value columns in this example; no missing-category behavior is asserted.
Stata reference · R reference 
hashsort g -x
 
data <- data[order(data$g,-data$x,method='radix'),,drop=FALSE]
base 		Stably sort by ascending group and descending value
Differs from Stata: Complete numeric keys only; R radix sort preserves original order for equal keys. No comparative speed claim is made.
Stata reference · R reference 
insobs 1, after(2)
 
data <- rbind(data[seq_len(2),,drop=FALSE],data[NA_integer_,,drop=FALSE],data[-seq_len(2),,drop=FALSE])
base 		Insert a missing row after a position
Fixed workflow example; adapt the object and column names together.
Stata reference · R reference 
insobs 1, before(3)
 
data <- rbind(data[1:2,,drop=FALSE],data[NA_integer_,,drop=FALSE],data[3:6,,drop=FALSE])
base 		Insert a missing row before a position
Fixed workflow example; adapt the object and column names together.
Stata reference · R reference 
insobs 2
 
data <- rbind(data,data[rep(NA_integer_,2),,drop=FALSE])
base 		Append two entirely missing observations
Fixed workflow example; adapt the object and column names together.
Stata reference · R reference 
inspect x
 
c(negative=sum(data$x<0,na.rm=TRUE),zero=sum(data$x==0,na.rm=TRUE),positive=sum(data$x>0,na.rm=TRUE),missing=sum(is.na(data$x)))
base 		Count signs and missing values
Differs from Stata: Sign/missing counts only; Stata's histogram and integer diagnostic are not included.
Stata reference · R reference 
ipolate y x, generate(z)
 
data$z <- stats::approx(data$x,data$y,xout=data$x,rule=1)$y
base 		Keep boundary missing values outside the interpolation range
Stata reference · R reference 
isid g h
 
stopifnot(!anyNA(data[c('g','h')]),!anyDuplicated(data[c('g','h')]))
base 		Assert a compound key uniquely identifies rows
Stata reference · R reference 
isid id
 
stopifnot(!anyNA(data$id),!anyDuplicated(data$id))
base 		Assert a key is unique and nonmissing
Stata reference · R reference 
isid id, missok
 
stopifnot(!anyDuplicated(data$id))
base 		Allow one ordinary missing value in a unique key
Stata reference · R reference 
joinby g using other.dta
 
data <- dplyr::inner_join(data,other,by='g',relationship='many-to-many')
dplyr 		Form every matching pair within groups
Stata reference · R reference 
joinby g using other.dta, unmatched(master)
 
data <- dplyr::left_join(data,other,by='g',relationship='many-to-many')
dplyr 		Keep unmatched master rows while forming group pairs
Differs from Stata: Returns paired rows without the Stata _merge indicator.
Stata reference · R reference 
keep id x
 
data <- data[c('id','x')]
base 		Keep named columns in order
Stata reference · R reference 
keep if !missing(x, y)
 
data <- data[stats::complete.cases(data[c('x','y')]), , drop=FALSE]
base 		Keep complete rows on selected columns
Stata reference · R reference 
keep if g == 2
 
data <- dplyr::filter(data, g == 2)
dplyr 		Filter to a category with dplyr
Stata reference · R reference 
keep if inlist(g, 1, 3)
 
data <- data[data$g %in% c(1,3), , drop=FALSE]
base 		Keep a set of categories
Stata reference · R reference 
keep if missing(x)
 
data <- data[is.na(data$x), , drop=FALSE]
base 		Keep rows missing a value
Stata reference · R reference 
keep if s != ""
 
data <- data[!is.na(data$s) & data$s != '', , drop=FALSE]
base 		Keep nonempty strings
Differs from Stata: Stata string missing is empty text; this excludes both empty text and R NA.
Stata reference · R reference 
keep if x > 2 & x < .
 
data <- data[!is.na(data$x) & data$x > 2, , drop=FALSE]
base 		Keep observed values above a threshold
Stata reference · R reference 
keep in 2/4
 
data <- data[2:4, , drop=FALSE]
base 		Keep a row interval by position
Fixed workflow example; adapt the object and column names together.
Stata reference · R reference 
levelsof g, local(groups)
 
groups <- sort(unique(data$g[!is.na(data$g)]))
base 		Store distinct observed numeric categories
Stata reference · R reference 
levelsof x, missing
 
sort(unique(data$x),na.last=TRUE)
base 		List distinct values including ordinary missing
Stata reference · R reference 
list id x if g == 2
 
data[data$g==2,c('id','x'),drop=FALSE]
base 		Inspect selected rows and columns
Stata reference · R reference 
list id x in 1/3
 
utils::head(data[c('id','x')],3)
utils 		Show selected columns from the first rows
Stata reference · R reference 
list in -2/l
 
utils::tail(data,2)
utils 		Show the last rows
Fixed workflow example; adapt the object and column names together.
Stata reference · R reference 
lookfor income
 
names(data)[grepl('income',names(data),ignore.case=TRUE)]
base 		Search column names without case sensitivity
Differs from Stata: Searches names only; include each column's label attribute to search variable labels too.
Stata reference · R reference 
memory
 
utils::object.size(data)
utils 		Measure the memory occupied by one R table
Reports one object's estimated memory, not Stata's process-wide memory allocation report.
Stata reference · R reference 
merge 1:1 id using other.dta, generate(status)
 
data$.master <- TRUE; other$.using <- TRUE; data <- dplyr::full_join(data,other,by='id',relationship='one-to-one'); data$status <- ifelse(is.na(data$.master),2L,ifelse(is.na(data$.using),1L,3L)); data$.master <- NULL; data$.using <- NULL
dplyr 		Record master-only, using-only and matched status
Requires disjoint non-key column names unless update is explicitly shown. dplyr suffixes overlapping names; ordinary Stata merge preserves master values, including master missing values, on matched rows.
Stata reference · R reference 
merge 1:1 id using other.dta, keep(master match) nogenerate
 
data <- dplyr::left_join(data,other,by='id',relationship='one-to-one')
dplyr 		Keep all master rows in a one-to-one merge
Requires disjoint non-key column names unless update is explicitly shown. dplyr suffixes overlapping names; ordinary Stata merge preserves master values, including master missing values, on matched rows.
Stata reference · R reference 
merge 1:1 id using other.dta, keep(match) nogenerate
 
data <- dplyr::inner_join(data,other,by='id',relationship='one-to-one')
dplyr 		Keep only matched keys in a one-to-one merge
Requires disjoint non-key column names unless update is explicitly shown. dplyr suffixes overlapping names; ordinary Stata merge preserves master values, including master missing values, on matched rows.
Stata reference · R reference 
merge 1:1 id using other.dta, keepusing(y) nogenerate
 
data <- dplyr::full_join(data,other[c('id','y')],by='id',relationship='one-to-one')
dplyr 		Limit using-table columns during merge
Requires disjoint non-key column names unless update is explicitly shown. dplyr suffixes overlapping names; ordinary Stata merge preserves master values, including master missing values, on matched rows.
Stata reference · R reference 
merge 1:1 id using other.dta, nogenerate
 
stopifnot(!anyDuplicated(data$id),!anyDuplicated(other$id)); data <- dplyr::full_join(data,other,by='id',relationship='one-to-one')
dplyr 		Perform a validated one-to-one full merge
Requires disjoint non-key column names unless update is explicitly shown. dplyr suffixes overlapping names; ordinary Stata merge preserves master values, including master missing values, on matched rows.
Stata reference · R reference 
merge 1:1 id using other.dta, update nogenerate
 
data <- dplyr::full_join(data,other,by='id',suffix=c('','.using'),relationship='one-to-one'); data$x <- dplyr::coalesce(data$x,data$x.using); data$x.using <- NULL
dplyr 		Fill master missing values from matched using values
Differs from Stata: Shows value update only; Stata update conflict status codes 4/5 are not generated. Requires disjoint non-key column names unless update is explicitly shown. dplyr suffixes overlapping names; ordinary Stata merge preserves master values, including master missing values, on matched rows.
Stata reference · R reference 
merge 1:m id using other.dta, nogenerate
 
data <- dplyr::full_join(data,other,by='id',relationship='one-to-many')
dplyr 		Expand unique master keys to multiple matching rows
Requires disjoint non-key column names unless update is explicitly shown. dplyr suffixes overlapping names; ordinary Stata merge preserves master values, including master missing values, on matched rows.
Stata reference · R reference 
merge m:1 g using groups.dta, keep(master match) nogenerate
 
data <- dplyr::left_join(data,groups,by='g',relationship='many-to-one')
dplyr 		Attach unique group attributes to many rows
Requires disjoint non-key column names unless update is explicitly shown. dplyr suffixes overlapping names; ordinary Stata merge preserves master values, including master missing values, on matched rows.
Stata reference · R reference 
missings dropobs x y, force
 
data <- data[rowSums(!is.na(data[c('x','y')]))>0,,drop=FALSE]
base 		Drop rows missing on every selected numeric variable
Stata reference · R reference 
missings dropvars, force
 
data <- data[!vapply(data,function(z) all(is.na(z)),logical(1))]
base 		Drop entirely missing numeric columns
Differs from Stata: Numeric-only fixture; Stata string missing is empty text and would require an additional R test for empty strings.
Stata reference · R reference 
missings report x y
 
colSums(is.na(data[c('x','y')]))
base 		Report missing-value counts by numeric variable
Differs from Stata: Numeric columns only. The R result includes zero counts; Stata report defaults to showing counts at least one.
Stata reference · R reference 
missings tag x y, generate(nmiss)
 
data$nmiss <- rowSums(is.na(data[c('x','y')]))
base 		Count numeric missing values within each row
Stata reference · R reference 
misstable patterns x y
 
table(apply(is.na(data[c('x','y')]),1,paste0,collapse=''))
base 		Count joint missingness patterns
Differs from Stata: TRUE marks missing in the R pattern string; Stata's displayed symbols/order differ.
Stata reference · R reference 
misstable summarize x y
 
colSums(is.na(data[c('x','y')]))
base 		Count missing values for each selected column
Stata reference · R reference 
mvdecode x, mv(-99 -98)
 
data$x[data$x %in% c(-99,-98)] <- NA_real_
base 		Convert multiple sentinel codes to missing
Stata reference · R reference 
mvdecode x, mv(-99=.a)
 
data$x[data$x %in% -99] <- haven::tagged_na('a')
haven 		Convert a sentinel to tagged missing
Differs from Stata: Tagged missing preserves the label but R comparisons still follow NA semantics.
Stata reference · R reference 
mvencode x, mv(-99)
 
data$x[is.na(data$x)] <- -99
base 		Encode missing values with a sentinel
Stata reference · R reference 
mvencode x, mv(.a=-99)
 
data$x[haven::is_tagged_na(data$x,'a')] <- -99
haven 		Encode one tagged missing category
Stata reference · R reference 
order _all, alphabetic
 
data <- data[sort(names(data))]
base 		Alphabetize column order
Fixed workflow example; adapt the object and column names together.
Stata reference · R reference 
order id, last
 
data <- dplyr::relocate(data, id, .after=dplyr::last_col())
dplyr 		Move a column to the end
Stata reference · R reference 
order s, after(id)
 
data <- dplyr::relocate(data, s, .after=id)
dplyr 		Move a column after another
Stata reference · R reference 
order y id, first
 
data <- data[c('y','id',setdiff(names(data),c('y','id')))]
base 		Move selected columns to the front
Stata reference · R reference 
putmata X=(x y)
 
X <- as.matrix(data[c('x','y')])
base 		Copy numeric columns into an R matrix
R matrix workflow replaces the separate Mata workspace; X is a fixed matrix name.
Stata reference · R reference 
putmata X=(x y), omitmissing
 
X <- as.matrix(data[stats::complete.cases(data[c('x','y')]),c('x','y')])
base 		Copy complete numeric rows into a matrix
R matrix workflow replaces Mata; rows are removed jointly over the selected columns.
Stata reference · R reference 
range z 0 1 5
 
data <- data.frame(z=seq(0,1,length.out=5))
base 		Create an evenly spaced grid including endpoints
Stata reference · R reference 
range z 0 10
 
data$z <- seq(0,10,length.out=nrow(data))
base 		Attach a grid using the current row count
Stata reference · R reference 
range z 10 0 6
 
data <- data.frame(z=seq(10,0,length.out=6))
base 		Create a decreasing numeric grid
Stata reference · R reference 
rangestat (mean) z=x, interval(key -1 1)
 
data$z <- vapply(data$key,function(k) mean(data$x[data$key>=k-1 & data$key<=k+1]),numeric(1))
base 		Compute a mean over each closed numeric-key neighborhood
Differs from Stata: Small complete-data workflow with inclusive key offsets and self included; quadratic scanning is not intended for large datasets.
Stata reference · R reference 
rangestat (mean) z=x, interval(key -1 1) excludeself
 
data$z <- vapply(seq_len(nrow(data)),function(i) mean(data$x[seq_len(nrow(data))!=i & data$key>=data$key[i]-1 & data$key<=data$key[i]+1]),numeric(1))
base 		Exclude the current row from numeric-range means
Differs from Stata: Excludes row identity rather than all tied keys. Every fixture neighborhood has another observation; empty neighborhoods need NA handling.
Stata reference · R reference 
rangestat (sum) z=x, interval(key -1 1)
 
data$z <- vapply(data$key,function(k) sum(data$x[data$key>=k-1 & data$key<=k+1]),numeric(1))
base 		Compute a sum over each closed numeric-key neighborhood
Differs from Stata: Complete finite keys/values only in this fixture; missing bounds, by-groups and weights need explicit extensions.
Stata reference · R reference 
recast double x
 
data$x <- as.double(data$x)
base 		Convert an integer column to double
Stata reference · R reference 
recast long x
 
data$x <- as.integer(data$x)
base 		Store whole-number values as R integers
Differs from Stata: Only use when values are whole numbers in R's integer range; Stata storage ranges differ.
Stata reference · R reference 
recode x (1/2=0) (3/6=1), generate(binary)
 
data$binary <- ifelse(is.na(data$x),NA_real_,ifelse(data$x<=2,0,1))
base 		Collapse integer categories to two groups
Differs from Stata: Fixture values are integers 1 through 6; specify additional branches for values outside those ranges.
Stata reference · R reference 
recode x (1=10) (2=20), generate(z)
 
data$z <- data$x; data$z[data$x %in% 1] <- 10; data$z[data$x %in% 2] <- 20
base 		Recode selected values and retain all others
Stata reference · R reference 
rename (x y) (income hours)
 
names(data)[match(c('x','y'),names(data))] <- c('income','hours')
base 		Rename several columns together
Stata reference · R reference 
rename *, lower
 
names(data) <- tolower(names(data))
base 		Lowercase all column names
Fixed workflow example; adapt the object and column names together.
Stata reference · R reference 
rename *, upper
 
names(data) <- toupper(names(data))
base 		Uppercase all column names
Fixed workflow example; adapt the object and column names together.
Stata reference · R reference 
rename old* new*
 
names(data) <- sub('^old','new',names(data))
base 		Replace a column-name prefix
Fixed workflow example; adapt the object and column names together.
Stata reference · R reference 
rename x income
 
data <- dplyr::rename(data, income=x)
dplyr 		Rename a column with dplyr
Stata reference · R reference 
replace s = "unknown" if s == ""
 
data$s[!is.na(data$s) & data$s==''] <- 'unknown'
base 		Replace empty text without replacing NA
Differs from Stata: R distinguishes empty text from NA; this example targets empty text only.
Stata reference · R reference 
replace x = . if x == 2
 
data$x[data$x %in% 2] <- NA_real_
base 		Convert a sentinel to missing
Stata reference · R reference 
replace x = 0 if missing(x)
 
data$x[is.na(data$x)] <- 0
base 		Replace numeric missing by zero
Stata reference · R reference 
replace x = x * 100 if g == 1
 
data$x[data$g %in% 1] <- data$x[data$g %in% 1] * 100
base 		Modify values only inside a category
Stata reference · R reference 
replace x = y if missing(x)
 
data$x[is.na(data$x)] <- data$y[is.na(data$x)]
base 		Fill missing values from another column
Stata reference · R reference 
reshape long x y, i(id) j(wave)
 
data <- tidyr::pivot_longer(data,cols=c(x1,x2,y1,y2),names_to=c('.value','wave'),names_pattern='([xy])([12])',names_transform=list(wave=as.integer))
tidyr 		Reshape multiple parallel measurement stubs
Fixed two-stub example; all name patterns and source names must be changed together.
Stata reference · R reference 
reshape long x, i(id) j(wave)
 
data <- tidyr::pivot_longer(data,cols=c(x1,x2),names_to='wave',names_prefix='x',names_transform=list(wave=as.integer),values_to='x')
tidyr 		Convert a wide measurement stub to long rows
Fixed x1/x2 stub example; source names, stub and output names must be changed together.
Stata reference · R reference 
reshape wide x, i(id) j(group) string
 
data <- tidyr::pivot_wider(data,id_cols=id,names_from=group,values_from=x,names_prefix='x')
tidyr 		Use text suffixes when widening measurements
Differs from Stata: Fixed string-suffix workflow; pair keys must be unique.
Stata reference · R reference 
reshape wide x, i(id) j(wave)
 
data <- tidyr::pivot_wider(data,id_cols=id,names_from=wave,values_from=x,names_prefix='x')
tidyr 		Convert unique long measurements to wide columns
Differs from Stata: Requires unique id-wave pairs; duplicates must be resolved explicitly rather than silently aggregated.
Stata reference · R reference 
separate x, by(g)
 
for (k in sort(unique(data$g))) data[[paste0('x',k)]] <- ifelse(data$g==k,data$x,NA_real_)
base 		Create group-specific copies of a numeric column
Stata reference · R reference 
separate y, by(g == 1) generate(z)
 
data$z0 <- ifelse(data$g!=1,data$y,NA_real_); data$z1 <- ifelse(data$g==1,data$y,NA_real_)
base 		Separate a column by a binary expression
The generated stub z is fixed here; z0 and z1 are the false and true branches.
Stata reference · R reference 
seqvar z = 1/3
 
data$z <- rep(NA_integer_,nrow(data)); data$z[seq_len(min(3L,nrow(data)))] <- seq_len(min(3L,nrow(data)))
base 		Leave remaining rows missing after a short integer sequence
Stata reference · R reference 
seqvar z = 1/3 5/7
 
data$z <- c(1:3,5:7)
base 		Create a specified integer sequence with a deliberate gap
Stata reference · R reference 
snapshot restore 1
 
data <- readRDS(path)
base 		Restore a temporary data snapshot
Restores an explicitly supplied RDS path rather than a numbered Stata snapshot.
Stata reference · R reference 
snapshot save
 
path <- tempfile(fileext='.rds'); saveRDS(data,path)
base 		Save a temporary on-disk data snapshot
Uses an RDS file path rather than Stata's numbered interactive snapshot manager.
Stata reference · R reference 
sort g x
 
data <- data[order(data$g,data$x,na.last=TRUE), , drop=FALSE]
base 		Sort lexicographically on two columns
Stata reference · R reference 
sort g, stable
 
data <- data[order(data$g,method='radix'), , drop=FALSE]
base 		Keep original order within equal keys
Stata reference · R reference 
sort x
 
data <- data[order(data$x, na.last=TRUE), , drop=FALSE]
base 		Sort a numeric column with missing last
Stata reference · R reference 
splitsample, generate(sample)
 
data$sample <- sample(rep(1:2,length.out=nrow(data)))
base 		Assign a balanced random half-split
Differs from Stata: Random assignments differ across R and Stata; this fixture has even sample size.
Stata reference · R reference 
splitsample, generate(sample) nsplit(3)
 
data$sample <- sample(rep(1:3,each=nrow(data)/3))
base 		Assign three equal random subsamples
Differs from Stata: Requires the fixture row count to be divisible by three; random assignments differ across systems.
Stata reference · R reference 
stack x a y b, into(value amount) clear
 
data <- rbind(data.frame(value=data$x,amount=data$a,`_stack`=1L,check.names=FALSE),data.frame(value=data$y,amount=data$b,`_stack`=2L,check.names=FALSE))
base 		Stack pairs of columns together
Stata reference · R reference 
stack x y, into(value) clear
 
data <- data.frame(value=c(data$x,data$y),`_stack`=rep(1:2,each=nrow(data)),check.names=FALSE)
base 		Stack two columns vertically with a source indicator
Stata reference · R reference 
tostring x, generate(s) format(%06.0f)
 
data$s <- sprintf('%06.0f',data$x)
base 		Format numeric identifiers with leading zeros
Stata reference · R reference 
tostring x, generate(s) format(%9.2f)
 
data$s <- sprintf('%.2f',data$x)
base 		Format numeric text to two decimal places
Differs from Stata: R example omits field-width padding; values are rounded for display and stored as text.
Stata reference · R reference 
xi i.g
 
data <- dplyr::select(tidystata::as_dummy(data, g, .names = "index"), -tidyselect::any_of("g_1"))
tidystata 		Create indicator variables omitting the lowest category
Differs from Stata: Keeps the original columns and omits the first observed category. Numeric categories are ordered by value; R factors follow their level order, which must match Stata's omitted category. Stata names indicators by category code; R numbers observed categories with .names="index". Use factor(g) in models; xi: interactions are not reproduced.
Stata reference · R reference 
xpose, clear
 
data <- as.data.frame(t(as.matrix(data)))
base 		Transpose an entirely numeric table
Differs from Stata: Numeric-only input; R output column names differ from Stata's v1, v2 names.
Stata reference · R reference 
xpose, clear varname
 
oldnames <- names(data); data <- as.data.frame(t(as.matrix(data))); data[['_varname']] <- oldnames
base 		Transpose values while retaining original column names
Differs from Stata: Numeric-only input. Original names are copied into _varname; R transposed column names may differ.
Stata reference · R reference

Files and import/export

Stata R Purpose, qualifications and sources 
changeeol source.txt destination.txt, eol(unix)
 
writeBin(charToRaw(paste0(paste(readLines(src),collapse='\n'),'\n')),dst)
base 		Normalize a text file to Unix newlines
Fixed workflow example; adapt the object and column names together.
Stata reference · R reference 
changeeol source.txt destination.txt, eol(windows)
 
writeBin(charToRaw(paste0(paste(readLines(src),collapse='\r\n'),'\r\n')),dst)
base 		Write text using explicit Windows CRLF bytes
Fixed workflow example; adapt the object and column names together.
Stata reference · R reference 
checksum sample.txt
 
tools::md5sum(src)
tools 		Compute a file integrity digest
MD5 is a workflow alternative: Stata checksum uses POSIX CRC, so digest values are not comparable.
Stata reference · R reference 
checksum sample.txt, saving(sample.sum)
 
digest <- tools::md5sum(src); writeLines(unname(digest),dst)
tools 		Save a file digest in a sidecar text file
Writes an R MD5 sidecar, not a Stata .sum file; Stata cannot validate this format as its CRC sidecar.
Stata reference · R reference 
copy source.txt destination.txt
 
file.copy(src,dst,overwrite=FALSE)
base 		Copy a local file to a new path
src and dst are temporary paths in this runnable fixture; no network transfer is involved.
Stata reference · R reference 
copy source.txt destination.txt, replace
 
file.copy(src,dst,overwrite=TRUE)
base 		Overwrite a destination file with a local copy
Demonstrates overwrite on newly created temporary files only.
Stata reference · R reference 
dir "*.csv"
 
list.files(path,pattern='[.]csv$',full.names=FALSE)
base 		List CSV files in a directory
Fixed workflow example; adapt the object and column names together.
Stata reference · R reference 
dir, all
 
list.files(path,all.files=TRUE,no..=TRUE)
base 		List files including hidden dot files
Differs from Stata: R all.files includes dot files; hidden-file conventions vary by operating system.
Stata reference · R reference 
export delimited id x using sample.csv, novarnames replace
 
utils::write.table(data[c('id','x')],path,sep=',',row.names=FALSE,col.names=FALSE,quote=FALSE,na='')
utils 		Export CSV without variable names
Stata reference · R reference 
export delimited id x using sample.csv, replace
 
utils::write.csv(data[c('id','x')],path,row.names=FALSE,na='')
utils 		Export selected columns to CSV
Stata reference · R reference 
export delimited id x using sample.tsv, delimiters("\t") replace
 
utils::write.table(data[c('id','x')],path,sep='\t',row.names=FALSE,quote=FALSE,na='')
utils 		Export a tab-separated table
Stata reference · R reference 
export excel id x using sample.xlsx, firstrow(variables) replace
 
writexl::write_xlsx(data[c('id','x')],path)
writexl 		Export selected columns to an Excel workbook
Stata reference · R reference 
export sasxport5 sample.xpt, replace
 
haven::write_xpt(data,path,version=5,name='SAMPLE')
haven 		Export data in SAS XPORT version 5 format
Fixed workflow example; adapt the object and column names together.
Stata reference · R reference 
export spss using sample.sav, replace
 
haven::write_sav(data,path)
haven 		Export labelled values to an SPSS system file
Fixed workflow example; adapt the object and column names together.
Stata reference · R reference 
file open handle using sample.txt, read text
file read handle line
file close handle
 
con <- file(src,open='rt',encoding='UTF-8'); line <- readLines(con,n=1); close(con)
base 		Read one line through a text connection
Fixed workflow example; adapt the object and column names together.
Stata reference · R reference 
file open handle using sample.txt, write text
file write handle "hello" _n
file close handle
 
con <- file(path,open='wt',encoding='UTF-8'); writeLines('hello',con); close(con)
base 		Write and close a UTF-8 text connection
The example closes its own temporary connection explicitly.
Stata reference · R reference 
filefilter source.txt destination.txt, from("old") to("new")
 
writeLines(gsub('old','new',readLines(src),fixed=TRUE),dst,useBytes=TRUE)
base 		Replace literal text throughout a small file
Differs from Stata: Line-oriented small-file workflow; binary patterns and large-file streaming are not covered.
Stata reference · R reference 
filefilter source.txt destination.txt, from(\t) to(";")
 
writeLines(gsub('\t',';',readLines(src),fixed=TRUE),dst,useBytes=TRUE)
base 		Replace tab separators in a text file
Differs from Stata: Literal small-text-file replacement; quoted delimiters in tabular data need a parser instead.
Stata reference · R reference 
findfile sample.txt, path(folder)
 
found <- list.files(folder,pattern='^sample[.]txt$',full.names=TRUE); stopifnot(length(found)==1L); normalizePath(found,winslash='/')
base 		Find a named file inside a specified directory
Fixed workflow example; adapt the object and column names together.
Stata reference · R reference 
hexdump sample.bin
 
bytes <- readBin(src,'raw',n=file.info(src)$size); format(bytes)
base 		Inspect a file as hexadecimal byte values
Fixed workflow example; adapt the object and column names together.
Stata reference · R reference 
hexdump sample.bin, analyze tabulate
 
table(as.integer(readBin(src,'raw',n=file.info(src)$size)))
base 		Count byte values in a binary file
Fixed workflow example; adapt the object and column names together.
Stata reference · R reference 
import delimited using sample.csv, clear
 
data <- utils::read.csv(path,stringsAsFactors=FALSE)
utils 		Import a CSV file with a header
Fixed workflow example; adapt the object and column names together.
Stata reference · R reference 
import delimited using sample.csv, rowrange(3:4) varnames(nonames) clear
 
data <- utils::read.csv(path,header=FALSE,skip=2,nrows=2,col.names=c('id','x'))
utils 		Read a selected range of text rows
Differs from Stata: Row numbers refer to physical file rows including the header; R column names are fixed in this example.
Stata reference · R reference 
import delimited using sample.csv, stringcols(1) clear
 
data <- utils::read.csv(path,colClasses=c('character','numeric'))
utils 		Preserve zero-padded identifiers during import
Fixed workflow example; adapt the object and column names together.
Stata reference · R reference 
import delimited using sample.csv, varnames(nonames) clear
 
data <- utils::read.csv(path,header=FALSE,col.names=c('id','x'))
utils 		Import a CSV file without a header
Differs from Stata: R column names id and x are supplied explicitly; Stata would choose default names without further renaming.
Stata reference · R reference 
import delimited using sample.tsv, delimiters("\t") clear
 
data <- utils::read.delim(path,stringsAsFactors=FALSE)
utils 		Import tab-separated values
Fixed workflow example; adapt the object and column names together.
Stata reference · R reference 
import excel using sample.xlsx, describe
 
readxl::excel_sheets(path)
readxl 		List worksheet names before importing
Fixed workflow example; adapt the object and column names together.
Stata reference · R reference 
import excel using sample.xlsx, sheet("Survey") cellrange(A1:B2) firstrow clear
 
data <- readxl::read_excel(path,sheet='Survey',range='A1:B2')
readxl 		Import a bounded Excel cell rectangle
Fixed workflow example; adapt the object and column names together.
Stata reference · R reference 
import excel using sample.xlsx, sheet("Survey") firstrow clear
 
data <- readxl::read_excel(path,sheet='Survey')
readxl 		Read one named Excel worksheet
Fixed workflow example; adapt the object and column names together.
Stata reference · R reference 
import sasxport5 sample.xpt, clear
 
data <- haven::read_xpt(path)
haven 		Import a SAS XPORT version 5 file
Fixed workflow example; adapt the object and column names together.
Stata reference · R reference 
import spss using sample.sav, clear
 
data <- haven::read_sav(path)
haven 		Import an SPSS system file
Fixed workflow example; adapt the object and column names together.
Stata reference · R reference 
import delimited using sample.csv, clear
 
data <- utils::read.csv(path,na.strings=c('','.'),stringsAsFactors=FALSE)
utils 		Recognize blank and period missing markers on import
Fixed workflow example; adapt the object and column names together.
Stata reference · R reference 
infile id x using sample.txt, clear
 
data <- utils::read.table(path,col.names=c('id','x'),header=FALSE)
utils 		Import whitespace-separated numeric fields
Free-format example with simple unquoted numeric fields.
Stata reference · R reference 
infile str8 name x using sample.txt, clear
 
data <- utils::read.table(path,col.names=c('name','x'),header=FALSE,stringsAsFactors=FALSE)
utils 		Import quoted text and numeric fields
Stata reference · R reference 
infix id 1-2 x 3-5 using sample.txt, clear
 
data <- utils::read.fwf(path,widths=c(2,3),col.names=c('id','x'))
utils 		Read adjacent fixed-width numeric fields
Stata reference · R reference 
infix str3 code 1-3 x 5-6 using sample.txt, clear
 
data <- utils::read.fwf(path,widths=c(3,-1,2),col.names=c('code','x'),stringsAsFactors=FALSE)
utils 		Read fixed-width fields while skipping a spacer
Stata reference · R reference 
jdbc insert id score, table("people")
 
DBI::dbAppendTable(con,'people',data); observed <- DBI::dbReadTable(con,'people'); DBI::dbDisconnect(con)
DBI 		Append controlled rows to an existing SQL table
Requires an existing Stata DSN/JDBC connection for the shown Stata example. The controlled SQLite fixture verifies the SQL/table operation, not ODBC/JDBC driver connection; use an appropriate DBI backend in production.
Stata reference · R reference 
jdbc load, table("people")
 
data <- DBI::dbReadTable(con,'people'); DBI::dbDisconnect(con)
DBI 		Load an entire SQL table through a DBI backend
Requires an existing Stata DSN/JDBC connection for the shown Stata example. The controlled SQLite fixture verifies the SQL/table operation, not ODBC/JDBC driver connection; use an appropriate DBI backend in production.
Stata reference · R reference 
mkdir work
 
dir.create(path)
base 		Create a directory
The runnable fixture creates a new directory under R's temporary directory.
Stata reference · R reference 
odbc describe "people", dsn("research")
 
fields <- DBI::dbListFields(con,'people'); DBI::dbDisconnect(con)
DBI 		Inspect database table field names
Requires an existing Stata DSN/JDBC connection for the shown Stata example. The controlled SQLite fixture verifies the SQL/table operation, not ODBC/JDBC driver connection; use an appropriate DBI backend in production. This row returns names only, not the Stata column-type report.
Stata reference · R reference 
odbc load, exec("SELECT id, score FROM people WHERE score >= 20") dsn("research")
 
data <- DBI::dbGetQuery(con,'SELECT id, score FROM people WHERE score >= 20 ORDER BY id'); DBI::dbDisconnect(con)
DBI 		Load selected rows from a SQL query
Requires an existing Stata DSN/JDBC connection for the shown Stata example. The controlled SQLite fixture verifies the SQL/table operation, not ODBC/JDBC driver connection; use an appropriate DBI backend in production.
Stata reference · R reference 
odbc query "research"
 
tables <- DBI::dbListTables(con); DBI::dbDisconnect(con)
DBI 		Inspect available tables in a database
Requires an existing Stata DSN/JDBC connection for the shown Stata example. The controlled SQLite fixture verifies the SQL/table operation, not ODBC/JDBC driver connection; use an appropriate DBI backend in production.
Stata reference · R reference 
outfile id x using sample.csv, comma replace
 
utils::write.table(data[c('id','x')],path,sep=',',row.names=FALSE,col.names=FALSE,quote=FALSE,na='.')
utils 		Export comma-separated fields without a header
Stata reference · R reference 
outfile id x using sample.txt, replace
 
utils::write.table(data[c('id','x')],path,sep=' ',row.names=FALSE,col.names=FALSE,quote=FALSE,na='.')
utils 		Export plain numeric fields with Stata-style missing text
Stata reference · R reference 
save sample.dta, replace
 
haven::write_dta(data,path)
haven 		Save a table as a current Stata DTA file
Differs from Stata: R object names are not stored as Stata's active-data state; haven writes supported Stata metadata.
Stata reference · R reference 
saveold sample.dta, version(12) replace
 
haven::write_dta(data,path,version=12)
haven 		Write a Stata 12-compatible DTA file
Differs from Stata: Use names accepted by haven's older-format validation; this fixture uses names of at least two characters.
Stata reference · R reference 
savespss "data.sav", replace
 
haven::write_sav(data, path)
haven 		Save the data as an SPSS file
Differs from Stata: Community command. haven writes variable and value labels, but rejects Stata tagged missing values (.a, .b) with an error. Convert them explicitly to ordinary NA or to chosen SPSS user-missing codes with haven::labelled_spss() before writing. String widths and display formats may differ.
Stata reference · R reference 
setroot
 
root <- rprojroot::find_root(rprojroot::has_file('.here'),path=child)
rprojroot 		Locate a project root from a marker file
Uses the explicit .here marker and a supplied starting path; the Stata command also supports other markers and stores a global macro.
Stata reference · R reference 
type sample.txt
 
readLines(src,encoding='UTF-8')
base 		Read lines from a UTF-8 text file
Differs from Stata: Returns text lines; print or cat them for a console display.
Stata reference · R reference 
type sample.txt, lines(1)
 
readLines(src,n=1,encoding='UTF-8')
base 		Read only the beginning of a text file
Fixed workflow example; adapt the object and column names together.
Stata reference · R reference 
unicode convertfile source.txt destination.txt, srcencoding("latin1") dstencoding("utf-8")
 
text <- readLines(src,encoding='latin1'); writeBin(charToRaw(paste0(iconv(text,from='latin1',to='UTF-8'),'\n')),dst)
base 		Convert a Latin-1 text file to UTF-8
Fixed workflow example; adapt the object and column names together.
Stata reference · R reference 
unzipfile sample.zip
 
utils::unzip(archive,exdir=destination)
utils 		Extract a ZIP archive into a chosen temporary directory
Differs from Stata: R extraction directory is explicit; Stata defaults to the current working directory.
Stata reference · R reference 
use id x using sample.dta, clear
 
data <- haven::read_dta(path,col_select=c('id','x'))
haven 		Read selected columns from a DTA file
Stata reference · R reference 
use sample.dta in 1/3, clear
 
data <- haven::read_dta(path,n_max=3)
haven 		Read the first observations from a DTA file
Fixed workflow example; adapt the object and column names together.
Stata reference · R reference 
zipfile sample.txt, saving(sample.zip)
 
zip::zipr(archive,src)
zip 		Create a ZIP archive containing a text file
Fixed workflow example; adapt the object and column names together.
Stata reference · R reference

Labels and metadata

Stata R Purpose, qualifications and sources 
char list x[]
 
attributes(data$x)
base 		Inspect custom column characteristics
Returns all R attributes, including any custom metadata.
Stata reference · R reference 
char x[source] "questionnaire"
 
attr(data$x,'source') <- 'questionnaire'
base 		Store a custom column characteristic
R attributes serve as local metadata; many transformations do not preserve arbitrary attributes.
Stata reference · R reference 
decode x, generate(text)
 
data$text <- as.character(haven::as_factor(data$x))
haven 		Decode numeric value labels to text
Stata reference · R reference 
encode s, generate(code)
 
data$code <- as.integer(factor(data$s,levels=sort(unique(data$s[!is.na(data$s)&data$s!='']))))
base 		Encode nonempty categories in sorted order
Differs from Stata: Creates integer codes only. Add a labelled vector or keep the factor to retain category labels; collation is locale-dependent.
Stata reference · R reference 
label data "Research sample"
 
attr(data,'label') <- 'Research sample'
base 		Attach a dataset label
Fixed workflow example; adapt the object and column names together.
Stata reference · R reference 
label define yesno 1 "No" 2 "Yes"
 
yesno <- c(No=1,Yes=2)
base 		Define a reusable value-label mapping
R stores the mapping as a named vector; attach it separately to each labelled column.
Stata reference · R reference 
label list yesno
 
attr(data$x,'labels')
base 		Read a column's value-label mapping
Fixed workflow example; adapt the object and column names together.
Stata reference · R reference 
label values x
 
data$x <- haven::zap_labels(data$x)
haven 		Remove value labels while preserving numeric values
Stata reference · R reference 
label values x yesno
 
data$x <- haven::labelled(data$x,labels=yesno)
haven 		Attach a numeric value-label dictionary
Stata reference · R reference 
label variable x ""
 
attr(data$x,'label') <- NULL
base 		Remove a variable label
Stata reference · R reference 
label variable x "Response"
 
attr(data$x,'label') <- 'Response'
base 		Attach a descriptive variable label
Stata reference · R reference 
labelbook
 
lapply(data,function(z) attr(z,'labels'))
base 		Inspect value-label dictionaries by column
Differs from Stata: Lists attached dictionaries; does not reproduce Stata's full multilingual label audit.
Stata reference · R reference 
labmask code, values(text)
 
dictionary <- unique(data[c('code','text')]); stopifnot(!anyDuplicated(dictionary$code)); data$code <- haven::labelled(data$code,stats::setNames(dictionary$code,dictionary$text))
haven 		Assign existing text values as numeric value labels
Differs from Stata: Requires a unique text label per observed numeric code; this preserves supplied numeric codes rather than alphabetically encoding text.
Stata reference · R reference 
notes list x
 
attr(data$x,'notes')
base 		Read notes attached to a column
Reads a custom R notes attribute; preservation depends on subsequent transformations.
Stata reference · R reference 
notes x: Measured at baseline
 
attr(data$x,'notes') <- 'Measured at baseline'
base 		Attach a text note to a column
R attribute workflow; numbered Stata note management is not recreated.
Stata reference · R reference 
numlabel, add
 
labs <- attr(data$x,'labels'); names(labs) <- paste0(unname(labs),'. ',names(labs)); attr(data$x,'labels') <- labs
base 		Prefix each label with its numeric code
Fixed workflow example; adapt the object and column names together.
Stata reference · R reference 
numlabel, remove
 
labs <- attr(data$x,'labels'); names(labs) <- sub('^[0-9]+[.] ','',names(labs)); attr(data$x,'labels') <- labs
base 		Remove a known numeric-code label prefix
Differs from Stata: Targets the demonstrated positive-integer prefix format; other masks require a matching pattern.
Stata reference · R reference 
uselabel using sample.dta, clear
 
labs <- attr(data$x,'labels'); dictionary <- data.frame(lname='x',value=unname(labs),label=names(labs),trunc=0L)
base 		Make a value-label dictionary table
Uses an already imported labelled column. R has no separate Stata label-definition name; lname uses the column name.
Stata reference · R reference

Descriptive tables

Stata R Purpose, qualifications and sources 
collapse (count) z=x, by(g)
 
data <- dplyr::summarise(data,z=sum(!is.na(x)),.by=g)
dplyr 		Replace rows with group nonmissing counts
Stata reference · R reference 
collapse (first) z=x, by(g)
 
data <- dplyr::summarise(data,z=dplyr::first(x),.by=g)
dplyr 		Replace rows with group first values including missing
Stata reference · R reference 
collapse (firstnm) z=x, by(g)
 
data <- dplyr::summarise(data,z=dplyr::first(x[!is.na(x)]),.by=g)
dplyr 		Replace rows with group first nonmissing values
Stata reference · R reference 
collapse (last) z=x, by(g)
 
data <- dplyr::summarise(data,z=dplyr::last(x),.by=g)
dplyr 		Replace rows with group last values including missing
Stata reference · R reference 
collapse (lastnm) z=x, by(g)
 
data <- dplyr::summarise(data,z=dplyr::last(x[!is.na(x)]),.by=g)
dplyr 		Replace rows with group last nonmissing values
Stata reference · R reference 
collapse (max) z=x, by(g)
 
data <- dplyr::summarise(data, z = if (all(is.na(x))) NA_real_ else max(x, na.rm = TRUE), .by = g)
dplyr 		Replace rows with group maxima
Stata reference · R reference 
collapse (mean) x [fw=w]
 
data <- data.frame(x=stats::weighted.mean(data$x,data$w))
stats 		Collapse integer-frequency-weighted values to a mean
Differs from Stata: Frequency-weighted mean only; no inference or weighted-variance formula is implied.
Stata reference · R reference 
collapse (mean) x y, cw by(g)
 
data <- data[stats::complete.cases(data[c('x','y')]),]; data <- dplyr::summarise(data,dplyr::across(c(x,y),mean),.by=g)
dplyr 		Use common complete cases for group summaries
Stata reference · R reference 
collapse (mean) z=x, by(g)
 
data <- dplyr::summarise(data,z=mean(x,na.rm=TRUE),.by=g)
dplyr 		Replace rows with group means
Stata reference · R reference 
collapse (median) z=x, by(g)
 
data <- dplyr::summarise(data,z=stats::median(x,na.rm=TRUE),.by=g)
dplyr 		Replace rows with group medians
Stata reference · R reference 
collapse (min) z=x, by(g)
 
data <- dplyr::summarise(data, z = if (all(is.na(x))) NA_real_ else min(x, na.rm = TRUE), .by = g)
dplyr 		Replace rows with group minima
Stata reference · R reference 
collapse (sum) z=x, by(g)
 
data <- dplyr::summarise(data,z=sum(x,na.rm=TRUE),.by=g)
dplyr 		Replace rows with group totals
Stata reference · R reference 
contract g [fw=w], freq(n)
 
data <- dplyr::summarise(data,n=sum(w),.by=g)
dplyr 		Create category counts from integer frequency weights
Stata reference · R reference 
contract g h, freq(n)
 
data <- dplyr::count(data,g,h,name='n')
dplyr 		Count observed combinations of two categories
Stata reference · R reference 
contract g h, freq(n) zero
 
data <- as.data.frame(table(data$g,data$h)); names(data) <- c('g','h','n')
base 		Include unobserved combinations with zero frequency
Stata reference · R reference 
contract g, freq(n)
 
data <- dplyr::count(data,g,name='n')
dplyr 		Replace observations with category frequencies
Stata reference · R reference 
contract g, freq(n) cfreq(cn) cpercent(cp)
 
data <- dplyr::count(data,g,name='n'); data$cn <- cumsum(data$n); data$cp <- 100*data$cn/sum(data$n)
dplyr 		Create cumulative frequencies and percentages
Stata reference · R reference 
contract g, freq(n) percent(p)
 
data <- dplyr::count(data,g,name='n'); data$p <- 100*data$n/sum(data$n)
dplyr 		Create category frequencies and percentages
Stata reference · R reference 
contract g, nomiss
 
data <- dplyr::count(data[!is.na(data$g)&data$g!='',,drop=FALSE],g,name='_freq')
dplyr 		Exclude missing categories before contracting
Differs from Stata: Excludes both empty strings and R NA to match Stata's nonmissing string categories.
Stata reference · R reference 
summarize x, detail
 
stats::quantile(data$x,c(.01,.05,.1,.25,.5,.75,.9,.95,.99),na.rm=TRUE,type=2)
base 		Get detailed empirical percentiles
Differs from Stata: Percentile portion of summarize detail, using R type 2; skewness/kurtosis are not returned here.
Stata reference · R reference 
summarize x, meanonly
 
c(n=sum(!is.na(data$x)),mean=mean(data$x,na.rm=TRUE),min=min(data$x,na.rm=TRUE),max=max(data$x,na.rm=TRUE))
base 		Compute mean-only summary components
Stata reference · R reference 
tab1 g h
 
lapply(data[c('g','h')],table)
base 		Make separate one-way frequency tables
Stata reference · R reference 
tab2 g h k
 
pairs <- utils::combn(c('g','h','k'),2,simplify=FALSE); lapply(pairs,function(p) table(data[[p[1]]],data[[p[2]]]))
base 		Make each pairwise contingency table
Stata reference · R reference 
tabdisp g h, cellvar(value)
 
stopifnot(!anyDuplicated(data[c('g','h')])); stats::xtabs(value~g+h,data=data,na.action=stats::na.pass)
stats 		Display a complete table of precomputed cell values
Differs from Stata: Requires one row for every cell and unique keys; xtabs would otherwise sum duplicates or fill absent cells with zero.
Stata reference · R reference 
tabstat x y, statistics(n mean sd) columns(statistics)
 
sapply(data[c('x','y')],function(z) c(n=sum(!is.na(z)),mean=mean(z,na.rm=TRUE),sd=stats::sd(z,na.rm=TRUE)))
base 		Compare count, mean and spread across numeric columns
Stata reference · R reference 
tabstat x, by(g) statistics(n mean)
 
data |> dplyr::group_by(g) |> dplyr::summarise(n=sum(!is.na(x)),mean=mean(x,na.rm=TRUE),.groups='drop')
dplyr 		Summarize nonmissing counts and means by category
Stata reference · R reference 
tabstat x, statistics(iqr)
 
with(data,diff(stats::quantile(x,c(.25,.75),na.rm=TRUE,type=2)))
base 		Return the interquartile range as a descriptive statistic
Differs from Stata: Uses the Stata-default empirical percentile rule through R quantile type 2.
Stata reference · R reference 
tabstat x, statistics(max)
 
with(data,max(x,na.rm=TRUE))
base 		Return the maximum as a descriptive statistic
Differs from Stata: Unweighted descriptive statistic; no survey or sampling-design correction is applied.
Stata reference · R reference 
tabstat x, statistics(mean)
 
with(data,mean(x,na.rm=TRUE))
base 		Return the mean as a descriptive statistic
Differs from Stata: Unweighted descriptive statistic; no survey or sampling-design correction is applied.
Stata reference · R reference 
tabstat x, statistics(min)
 
with(data,min(x,na.rm=TRUE))
base 		Return the minimum as a descriptive statistic
Differs from Stata: Unweighted descriptive statistic; no survey or sampling-design correction is applied.
Stata reference · R reference 
tabstat x, statistics(n)
 
with(data,sum(!is.na(x)))
base 		Return the nonmissing count as a descriptive statistic
Differs from Stata: Unweighted descriptive statistic; no survey or sampling-design correction is applied.
Stata reference · R reference 
tabstat x, statistics(p25)
 
with(data,as.numeric(stats::quantile(x,.25,na.rm=TRUE,type=2)))
base 		Return the 25th percentile as a descriptive statistic
Differs from Stata: Uses the Stata-default empirical percentile rule through R quantile type 2.
Stata reference · R reference 
tabstat x, statistics(p50)
 
with(data,stats::median(x,na.rm=TRUE))
base 		Return the median as a descriptive statistic
Differs from Stata: Unweighted descriptive statistic; no survey or sampling-design correction is applied.
Stata reference · R reference 
tabstat x, statistics(p75)
 
with(data,as.numeric(stats::quantile(x,.75,na.rm=TRUE,type=2)))
base 		Return the 75th percentile as a descriptive statistic
Differs from Stata: Uses the Stata-default empirical percentile rule through R quantile type 2.
Stata reference · R reference 
tabstat x, statistics(range)
 
with(data,diff(range(x,na.rm=TRUE)))
base 		Return the range as a descriptive statistic
Differs from Stata: Unweighted descriptive statistic; no survey or sampling-design correction is applied.
Stata reference · R reference 
tabstat x, statistics(sd)
 
with(data,stats::sd(x,na.rm=TRUE))
base 		Return the sample standard deviation as a descriptive statistic
Differs from Stata: Unweighted descriptive statistic; no survey or sampling-design correction is applied.
Stata reference · R reference 
tabstat x, statistics(semean)
 
with(data,stats::sd(x,na.rm=TRUE)/sqrt(sum(!is.na(x))))
base 		Return the standard error of the mean as a descriptive statistic
Differs from Stata: Unweighted descriptive statistic; no survey or sampling-design correction is applied.
Stata reference · R reference 
tabstat x, statistics(sum)
 
with(data,sum(x,na.rm=TRUE))
base 		Return the sum as a descriptive statistic
Differs from Stata: Unweighted descriptive statistic; no survey or sampling-design correction is applied.
Stata reference · R reference 
tabstat x, statistics(variance)
 
with(data,stats::var(x,na.rm=TRUE))
base 		Return the sample variance as a descriptive statistic
Differs from Stata: Unweighted descriptive statistic; no survey or sampling-design correction is applied.
Stata reference · R reference 
tabulate g
 
table(data$g)
base 		Count a single categorical variable
Stata reference · R reference 
tabulate g [fw=w]
 
stats::xtabs(w~g,data=data,na.action=stats::na.pass)
stats 		Tabulate integer frequency weights
Stata reference · R reference 
tabulate g h
 
table(data$g,data$h)
base 		Count a two-way contingency table
Stata reference · R reference 
tabulate g h, cell
 
100*prop.table(table(data$g,data$h))
base 		Compute percentages of all table cells
Stata reference · R reference 
tabulate g h, column
 
100*prop.table(table(data$g,data$h),margin=2)
base 		Compute percentages within each table column
Stata reference · R reference 
tabulate g h, row
 
100*prop.table(table(data$g,data$h),margin=1)
base 		Compute percentages within each table row
Stata reference · R reference 
tabulate g, generate(d)
 
indicators <- tidystata::as_dummy(dplyr::select(data, g), g, .remove = TRUE)
tidystata 		Create a full set of category indicators
Differs from Stata: Keeps all observations; a missing category gives NA in every dummy. R returns a separate data frame with category-based names; the Stata stub d is not a templated output name.
Stata reference · R reference 
tabulate g, missing
 
table(data$g,useNA='ifany')
base 		Include missing categories in a frequency table
Differs from Stata: Maps R NA to an explicit table category; empty character strings remain a separate R value.
Stata reference · R reference

Models and tests

Stata R Purpose, qualifications and sources 
ameans y
 
c(arithmetic=mean(data$y),geometric=exp(mean(log(data$y))),harmonic=1/mean(1/data$y))
stats 		Compare arithmetic, geometric and harmonic means
Differs from Stata: Returns the three point estimates for positive nonmissing values; Stata also reports their confidence intervals.
Stata reference · R reference 
biprobit y1 y2 x
 
VGAM::vglm(cbind(y1,y2)~x,family=VGAM::binom2.rho(),data=data)
VGAM 		Fit jointly correlated binary probit equations
Differs from Stata: Both binary equations share x and use a constant latent correlation; rho link parameterizations differ but fitted joint probabilities have the same model meaning.
Stata reference · R reference 
centile y, centile(50)
 
{ fit <- boot::boot(data,function(d,i) unname(stats::quantile(d$y[i],0.5,type=6)),R=499); list(median=fit$t0,interval=boot::boot.ci(fit,type="perc")$percent[4:5]) }
boot 		Estimate a median and bootstrap its confidence interval
Point-estimate correspondence; this uses a percentile bootstrap interval, whereas Stata centile defaults to a binomial order-statistic interval.
Stata reference · R reference 
contrast r.g
 
multcomp::glht(fit,linfct=multcomp::mcp(g="Dunnett"))
multcomp 		Form group contrasts against the reference category
Differs from Stata: Returns contrast estimates and covariance. Stata contrast defaults to unadjusted tests; summary(glht) defaults to single-step multiplicity adjustment. Use summary(result, test=multcomp::adjusted("none")) for unadjusted p-values.
Stata reference · R reference 
cpoisson y x, ul(4)
 
VGAM::vglm(VGAM::SurvS4(y,status)~x,family=VGAM::cens.poisson(),data=data)
VGAM 		Fit a right-censored Poisson count model
Differs from Stata: Fixed right-censoring limit 4; status is one for exact counts and zero for counts known to be at least 4. Censoring must not be treated as truncation.
Stata reference · R reference 
elasticnet linear y x z, alpha(0.5)
 
glmnet::cv.glmnet(as.matrix(data[c("x","z")]),data$y,alpha=0.5,nfolds=10)
glmnet 		Fit cross-validated elasticnet linear prediction
Prediction workflow: folds, standardization, penalty grids, loss scaling and lambda selection differ from Stata defaults; coefficients need not match.
Stata reference · R reference 
esize twosample y, by(g) cohensd
 
effectsize::cohens_d(y~g,data=data,pooled_sd=TRUE)
effectsize 		Estimate Cohen standardized mean differences for two independent groups
Differs from Stata: Uses pooled SD and first-group minus second-group direction; reverse the sign if the substantive contrast is reversed.
Stata reference · R reference 
esizei 30 10 2 40 12 3, cohensd
 
(10-12)/sqrt(((30-1)*2^2+(40-1)*3^2)/(30+40-2))
stats 		Compute a pooled-SD effect size from group summaries
Fixed summary-statistic workflow returns Cohen d only; confidence intervals and Hedges correction are not included.
Stata reference · R reference 
icc y id
 
{ wide <- stats::reshape(data,idvar="id",timevar="rater",direction="wide"); psych::ICC(wide[,-1],lmer=FALSE)$results[1,,drop=FALSE] }
psych 		Estimate a one-way individual-measure intraclass correlation
Differs from Stata: Fixed balanced design has three measurements per target. R starts from wide repeated-measure columns; ICC1 is the one-way absolute-agreement individual-measure form.
Stata reference · R reference 
lasso linear y x z
 
glmnet::cv.glmnet(as.matrix(data[c("x","z")]),data$y,alpha=1,nfolds=10)
glmnet 		Fit cross-validated lasso linear prediction
Prediction workflow: folds, standardization, penalty grids, loss scaling and lambda selection differ from Stata defaults; coefficients need not match.
Stata reference · R reference 
manovatest g
 
summary(fit,test="Pillai")$stats["g",,drop=FALSE]
stats 		Test a factor term in a fitted multivariate model
Differs from Stata: After fitting the corresponding MANOVA model. This R row extracts Pillai; Stata manovatest reports four multivariate statistics.
Stata reference · R reference 
nestreg: regress y (x) (z)
 
{ reduced <- stats::lm(y~x,data=data); full <- stats::lm(y~x+z,data=data); stats::anova(reduced,full) }
stats 		Compare successively added regression blocks
Differs from Stata: Compares these two nested OLS fits with a partial F test; it does not implement the general Stata prefix that sequences arbitrary model blocks.
Stata reference · R reference 
orthog x z, generate(ox oz)
 
{ Q <- qr.Q(qr(cbind(1,as.matrix(data[c("x","z")])))); data[c("ox","oz")] <- Q[,-1,drop=FALSE]*sqrt(nrow(data)); data[c("ox","oz")] }
stats 		Orthogonalize centered predictor columns in their stated order
Differs from Stata: Unweighted QR gives an orthogonal basis with mean square one; signs can differ from Stata modified Gram-Schmidt output.
Stata reference · R reference 
orthpoly x, generate(p1 p2 p3) degree(3)
 
data[c("p1","p2","p3")] <- stats::poly(data$x,degree=3)*sqrt(nrow(data))
stats 		Construct three orthogonal polynomial columns
Differs from Stata: Rescales the R orthonormal basis to mean square one; polynomial signs and stored coefficient conventions may differ.
Stata reference · R reference 
predictnl p = invlogit(predict(xb)), se(se)
 
{ pred <- stats::predict(fit,type="link",se.fit=TRUE); data$p <- stats::plogis(pred$fit); data$se <- data$p*(1-data$p)*pred$se.fit; data[c("p","se")] }
stats 		Apply a nonlinear prediction transform with delta-method standard errors
Differs from Stata: Run after a logit GLM. This uses the derivative p*(1-p) and the fitted linear-index standard error.
Stata reference · R reference 
pwcompare g, mcompare(tukey)
 
stats::TukeyHSD(stats::aov(y~g,data=data))$g
stats 		Compare all group pairs with Tukey simultaneous intervals
Differs from Stata: After the one-way equal-variance model y on g; R TukeyHSD is not a general replacement for adjusted margins in arbitrary models.
Stata reference · R reference 
pwmean y, over(g) mcompare(bonferroni)
 
stats::pairwise.t.test(data$y,data$g,p.adjust.method="bonferroni",pool.sd=TRUE)
stats 		Compare group means using Bonferroni-adjusted pooled tests
Differs from Stata: Uses a common pooled within-group variance and Bonferroni-adjusted pairwise t tests. The returned p-value matrix does not include Stata pairwise mean-difference confidence intervals.
Stata reference · R reference 
robvar y, by(g)
 
car::leveneTest(y~factor(g),data=data,center=stats::median)
car 		Test equal group variances using median-centered Levene inference
Differs from Stata: Corresponds to robvar W50; Stata additionally reports mean-centered W0 and 10-percent-trimmed W10.
Stata reference · R reference 
runtest y, threshold(0)
 
{ above <- data$y>0; n1 <- sum(above); n0 <- sum(!above); runs <- 1+sum(diff(above)!=0); expected <- 1+2*n1*n0/(n1+n0); variance <- 2*n1*n0*(2*n1*n0-n1-n0)/((n1+n0)^2*(n1+n0-1)); z <- (runs-expected)/sqrt(variance); c(runs=runs,z=z,p.value=2*stats::pnorm(-abs(z))) }
stats 		Test whether a binary threshold sequence has random runs
Differs from Stata: Uses the large-sample uncorrected runs statistic, matching Stata default continuity convention; missing values must be removed before this workflow.
Stata reference · R reference 
slogit y x z, dimension(1) baseoutcome(3)
 
VGAM::rrvglm(y~x+z,family=VGAM::multinomial(refLevel="3"),data=data,Rank=1)
VGAM 		Fit a rank-one stereotype multinomial logit model
Differs from Stata: Uses the reduced-rank multinomial-logit form of the one-dimensional Anderson stereotype model. Category score identification, scale and sign conventions can differ.
Stata reference · R reference 
symmi 10 8 \ 3 10
 
stats::mcnemar.test(matrix(c(10,8,3,10),2,byrow=TRUE),correct=FALSE)
stats 		Test matched-table symmetry directly from cell counts
Differs from Stata: For this 2-by-2 table the symmetry test is the uncorrected McNemar test. Stata also reports table summaries; this R result is an htest object.
Stata reference · R reference 
tnbreg y x, ll(0)
 
VGAM::vglm(y~x,family=VGAM::posnegbinomial(),data=data)
VGAM 		Fit zero-truncated NB2 regression
Differs from Stata: Zero-truncated NB2 with constant size; VGAM size equals 1/alpha. Information-matrix and optimizer conventions can differ.
Stata reference · R reference 
tpoisson y x, ll(0)
 
VGAM::vglm(y~x,family=VGAM::pospoisson(),data=data)
VGAM 		Fit zero-truncated Poisson regression
Differs from Stata: Zero-truncation only: fitted values are conditional positive-count means, while log-link coefficients parameterize the latent Poisson lambda.
Stata reference · R reference 
truncreg y x, ll(0)
 
VGAM::vglm(y~x,family=VGAM::posnormal(),data=data)
VGAM 		Fit a zero-left-truncated normal regression
Differs from Stata: VGAM posnormal represents the special case of normal observations selected above zero; this example does not cover varying or upper truncation limits.
Stata reference · R reference 
ztest y == 20, sd(3)
 
{ z <- (mean(data$y)-20)/(3/sqrt(nrow(data))); c(z=z,p.value=2*stats::pnorm(-abs(z))) }
stats 		Test a mean using a known population standard deviation
Differs from Stata: The value 3 is assumed known as a population SD, not estimated from the sample.
Stata reference · R reference 
ztesti 30 10 2 40 12 3
 
{ z <- (10-12)/sqrt(2^2/30+3^2/40); c(z=z,p.value=2*stats::pnorm(-abs(z))) }
stats 		Compare independent means from known-variance summaries
Differs from Stata: Treats the supplied standard deviations as known and uses a normal reference distribution. Returns the z statistic and two-sided p-value, without Stata confidence limits.
Stata reference · R reference

Statistical distributions

Stata R Purpose, qualifications and sources 
display F(4,10,2)
 
stats::pf(2, df1=4, df2=10)
stats 		Evaluate the F cumulative probability
Stata reference · R reference 
display Fden(4,10,2)
 
stats::df(2, df1=4, df2=10)
stats 		Evaluate the F density
Stata reference · R reference 
display Ftail(4,10,2)
 
stats::pf(2, df1=4, df2=10, lower.tail=FALSE)
stats 		Evaluate the F upper-tail probability
Stata reference · R reference 
display betaden(2,3,0.4)
 
stats::dbeta(0.4, shape1=2, shape2=3)
stats 		Evaluate the beta density
Stata reference · R reference 
display binomial(10,3,0.4)
 
stats::pbinom(3, size=10, prob=0.4)
stats 		Compute an inclusive binomial lower tail
Stata reference · R reference 
display binomialp(10,3,0.4)
 
stats::dbinom(3, size=10, prob=0.4)
stats 		Evaluate a binomial probability mass
Stata reference · R reference 
display binomialtail(10,3,0.4)
 
stats::pbinom(2, size=10, prob=0.4, lower.tail=FALSE)
stats 		Compute an inclusive binomial upper tail
Stata includes k in the upper tail; R lower.tail=FALSE excludes its q argument, so use k-1.
Stata reference · R reference 
display binormal(0,0,0.4)
 
as.numeric(mvtnorm::pmvnorm(upper=c(0,0),corr=matrix(c(1,0.4,0.4,1),2)))
mvtnorm 		Evaluate a correlated bivariate normal probability
Differs from Stata: Evaluates a bivariate normal probability using mvtnorm numerical integration. Integration tolerance and numerical rounding differ from the Stata distribution function.
Stata reference · R reference 
display cauchy(1,2,3)
 
stats::pcauchy(3, location=1, scale=2)
stats 		Evaluate the Cauchy cumulative probability
Stata reference · R reference 
display cauchyden(1,2,3)
 
stats::dcauchy(3, location=1, scale=2)
stats 		Evaluate the Cauchy density
Stata reference · R reference 
display cauchytail(1,2,3)
 
stats::pcauchy(3, location=1, scale=2, lower.tail=FALSE)
stats 		Evaluate the Cauchy upper-tail probability
Stata reference · R reference 
display chi2(4,3)
 
stats::pchisq(3, df=4)
stats 		Evaluate the chi-squared cumulative probability
Stata reference · R reference 
display chi2den(4,3)
 
stats::dchisq(3, df=4)
stats 		Evaluate the chi-squared density
Stata reference · R reference 
display chi2tail(4,3)
 
stats::pchisq(3, df=4, lower.tail=FALSE)
stats 		Evaluate the chi-squared upper-tail probability
Stata reference · R reference 
corr2data x y, n(100)
 
stats::setNames(as.data.frame(MASS::mvrnorm(n=100,mu=c(0,0),Sigma=diag(2),empirical=TRUE)),c("x","y"))
MASS 		Generate pseudo-data with exactly specified sample moments
Differs from Stata: empirical=TRUE imposes the sample mean and covariance; such pseudo-data are not independent random draws.
Stata reference · R reference 
drawnorm x y, n(100)
 
stats::setNames(as.data.frame(MASS::mvrnorm(n=100,mu=c(0,0),Sigma=diag(2))),c("x","y"))
MASS 		Draw independent standard normal columns jointly
Differs from Stata: Random streams differ; R returns a data frame of generated columns.
Stata reference · R reference 
display exponential(2,3)
 
stats::pexp(3, rate=1/2)
stats 		Evaluate the exponential cumulative probability
Stata reference · R reference 
display exponentialden(2,3)
 
stats::dexp(3, rate=1/2)
stats 		Evaluate the exponential density
Stata reference · R reference 
display exponentialtail(2,3)
 
stats::pexp(3, rate=1/2, lower.tail=FALSE)
stats 		Evaluate the exponential upper-tail probability
Stata reference · R reference 
display gammaden(3,1,0,2)
 
stats::dgamma(2, shape=3)
stats 		Evaluate the gamma density
Stata reference · R reference 
display gammaden(3,2,1,5)
 
stats::dgamma(5-1, shape=3, scale=2)
stats 		Evaluate a shifted gamma density
Stata reference · R reference 
display gammap(3,2)
 
stats::pgamma(2, shape=3)
stats 		Evaluate the gamma cumulative probability
Stata reference · R reference 
display gammaptail(3,2)
 
stats::pgamma(2, shape=3, lower.tail=FALSE)
stats 		Evaluate the gamma upper-tail probability
Stata reference · R reference 
display hypergeometric(20,8,5,3)
 
stats::phyper(3, m=8, n=12, k=5)
stats 		Compute an inclusive hypergeometric lower tail
Stata reference · R reference 
display hypergeometricp(20,8,5,3)
 
stats::dhyper(3, m=8, n=12, k=5)
stats 		Evaluate a hypergeometric probability mass
Stata reference · R reference 
display ibeta(2,3,0.4)
 
stats::pbeta(0.4, shape1=2, shape2=3)
stats 		Evaluate the beta cumulative probability
Stata reference · R reference 
display ibetatail(2,3,0.4)
 
stats::pbeta(0.4, shape1=2, shape2=3, lower.tail=FALSE)
stats 		Evaluate the beta upper-tail probability
Stata reference · R reference 
display invF(4,10,0.9)
 
stats::qf(0.9, df1=4, df2=10)
stats 		Find the F 90th percentile
Stata reference · R reference 
display invFtail(4,10,0.9)
 
stats::qf(0.9, df1=4, df2=10, lower.tail=FALSE)
stats 		Find a F upper-tail critical value
Stata reference · R reference 
display invbinomial(10,3,0.5)
 
stats::qbeta(0.5, shape1=4, shape2=7, lower.tail=FALSE)
stats 		Solve a binomial CDF for its success probability
These Stata inverse functions solve for a distribution parameter, not a discrete outcome quantile.
Stata reference · R reference 
display invbinomialtail(10,3,0.5)
 
stats::qbeta(0.5, shape1=3, shape2=8)
stats 		Solve an inclusive binomial tail for success probability
These Stata inverse functions solve for a distribution parameter, not a discrete outcome quantile.
Stata reference · R reference 
display invcauchy(1,2,0.9)
 
stats::qcauchy(0.9, location=1, scale=2)
stats 		Find the Cauchy 90th percentile
Stata reference · R reference 
display invcauchytail(1,2,0.9)
 
stats::qcauchy(0.9, location=1, scale=2, lower.tail=FALSE)
stats 		Find a Cauchy upper-tail critical value
Stata reference · R reference 
display invchi2(4,0.9)
 
stats::qchisq(0.9, df=4)
stats 		Find the chi-squared 90th percentile
Stata reference · R reference 
display invchi2tail(4,0.9)
 
stats::qchisq(0.9, df=4, lower.tail=FALSE)
stats 		Find a chi-squared upper-tail critical value
Stata reference · R reference 
display invexponential(2,0.9)
 
stats::qexp(0.9, rate=1/2)
stats 		Find the exponential 90th percentile
Stata reference · R reference 
display invexponentialtail(2,0.9)
 
stats::qexp(0.9, rate=1/2, lower.tail=FALSE)
stats 		Find a exponential upper-tail critical value
Stata reference · R reference 
display invgammap(3,0.9)
 
stats::qgamma(0.9, shape=3)
stats 		Find the gamma 90th percentile
Stata reference · R reference 
display invgammaptail(3,0.9)
 
stats::qgamma(0.9, shape=3, lower.tail=FALSE)
stats 		Find a gamma upper-tail critical value
Stata reference · R reference 
display invibeta(2,3,0.9)
 
stats::qbeta(0.9, shape1=2, shape2=3)
stats 		Find the beta 90th percentile
Stata reference · R reference 
display invibetatail(2,3,0.9)
 
stats::qbeta(0.9, shape1=2, shape2=3, lower.tail=FALSE)
stats 		Find a beta upper-tail critical value
Stata reference · R reference 
display invlogistic(0.8)
 
stats::qlogis(0.8)
stats 		Convert probability to log odds
Stata reference · R reference 
display invlogistic(1,2,0.9)
 
stats::qlogis(0.9, location=1, scale=2)
stats 		Find the logistic 90th percentile
Stata reference · R reference 
display invlogistictail(1,2,0.9)
 
stats::qlogis(0.9, location=1, scale=2, lower.tail=FALSE)
stats 		Find a logistic upper-tail critical value
Stata reference · R reference 
display invnF(4,20,2,0.9)
 
stats::qf(0.9, df1=4, df2=20, ncp=2)
stats 		Find a nF noncentral critical value
Stata reference · R reference 
display invnFtail(4,20,2,0.9)
 
stats::qf(0.9, df1=4, df2=20, ncp=2, lower.tail=FALSE)
stats 		Find a nF noncentral upper-tail critical value
Stata reference · R reference 
display invnbinomial(4,3,0.5)
 
stats::qbeta(0.5, shape1=4, shape2=4)
stats 		Solve a negative-binomial CDF for success probability
These Stata inverse functions solve for a distribution parameter, not a discrete outcome quantile.
Stata reference · R reference 
display invnbinomialtail(4,3,0.5)
 
stats::qbeta(0.5, shape1=4, shape2=3, lower.tail=FALSE)
stats 		Solve a negative-binomial tail for success probability
These Stata inverse functions solve for a distribution parameter, not a discrete outcome quantile.
Stata reference · R reference 
display invnchi2(4,2,0.9)
 
stats::qchisq(0.9, df=4, ncp=2)
stats 		Find a nchi2 noncentral critical value
Stata reference · R reference 
display invnchi2tail(4,2,0.9)
 
stats::qchisq(0.9, df=4, ncp=2, lower.tail=FALSE)
stats 		Find a nchi2 noncentral upper-tail critical value
Stata reference · R reference 
display invnibeta(2,3,2,0.9)
 
stats::qbeta(0.9, shape1=2, shape2=3, ncp=2)
stats 		Find a nibeta noncentral critical value
Stata reference · R reference 
display invnormal(0.9)
 
stats::qnorm(0.9)
stats 		Find the standard normal 90th percentile
Stata reference · R reference 
display invnt(20,2,0.9)
 
stats::qt(0.9, df=20, ncp=2)
stats 		Find a nt noncentral critical value
Stata reference · R reference 
display invnttail(20,2,0.9)
 
stats::qt(0.9, df=20, ncp=2, lower.tail=FALSE)
stats 		Find a nt noncentral upper-tail critical value
Stata reference · R reference 
display invpoisson(3,0.5)
 
stats::qgamma(0.5, shape=4, lower.tail=FALSE)
stats 		Solve a Poisson CDF for its mean
These Stata inverse functions solve for a distribution parameter, not a discrete outcome quantile.
Stata reference · R reference 
display invpoissontail(3,0.5)
 
stats::qgamma(0.5, shape=3)
stats 		Solve an inclusive Poisson tail for its mean
These Stata inverse functions solve for a distribution parameter, not a discrete outcome quantile.
Stata reference · R reference 
display invt(10,0.9)
 
stats::qt(0.9, df=10)
stats 		Find the Student t 90th percentile
Stata reference · R reference 
display invttail(10,0.9)
 
stats::qt(0.9, df=10, lower.tail=FALSE)
stats 		Find a Student t upper-tail critical value
Stata reference · R reference 
display invtukeyprob(4,20,0.95)
 
stats::qtukey(0.95, nmeans=4, df=20)
stats 		Find a Tukey simultaneous-comparison critical value
Differs from Stata: Uses the studentized-range quantile from qtukey. R and Stata implement different numerical approximations; tested inputs differed by about 2.5e-8, so compare within numerical tolerance rather than exact equality.
Stata reference · R reference 
display invweibull(2,3,0.9)
 
stats::qweibull(0.9, shape=2, scale=3)
stats 		Find the Weibull 90th percentile
Stata reference · R reference 
display invweibullph(2,0.5,0.9)
 
stats::qweibull(0.9, shape=2, scale=0.5^(-1/2))
stats 		Find a PH Weibull percentile
Stata reference · R reference 
display invweibullphtail(2,0.5,0.1)
 
stats::qweibull(0.1, shape=2, scale=0.5^(-1/2), lower.tail=FALSE)
stats 		Find a PH Weibull upper-tail quantile
Stata reference · R reference 
display invweibulltail(2,3,0.9)
 
stats::qweibull(0.9, shape=2, scale=3, lower.tail=FALSE)
stats 		Find a Weibull upper-tail critical value
Stata reference · R reference 
display lncauchyden(1,2,3)
 
stats::dcauchy(3, location=1, scale=2, log=TRUE)
stats 		Evaluate a Cauchy log density
Stata reference · R reference 
display lnigammaden(3,2,4)
 
stats::dgamma(1/4,shape=3,rate=2,log=TRUE)-2*log(4)
stats 		Evaluate an inverse-gamma log density
Differs from Stata: Uses the reciprocal-gamma change of variables with its log-Jacobian. Match Stata shape and scale parameterization; numerical tail underflow can differ.
Stata reference · R reference 
matrix M = (0,0)
matrix V = I(2)
matrix X = (1,2)
display lnmvnormalden(M,V,X)
 
mvtnorm::dmvnorm(x=c(1,2),mean=c(0,0),sigma=diag(2),log=TRUE)
mvtnorm 		Evaluate a multivariate normal log density
Differs from Stata: Fixed example uses mean M=(0,0), covariance V=I(2), and evaluation vector X=(1,2).
Stata reference · R reference 
display lnnormal(-8)
 
stats::pnorm(-8, log.p=TRUE)
stats 		Compute a stable extreme normal log probability
Stata reference · R reference 
display lnnormalden(3,1,2)
 
stats::dnorm(3, mean=1, sd=2, log=TRUE)
stats 		Evaluate a normal log density with location and scale
Stata reference · R reference 
display logistic(0.7)
 
stats::plogis(0.7)
stats 		Convert log odds to probability
Stata reference · R reference 
display logistic(1,2,3)
 
stats::plogis(3, location=1, scale=2)
stats 		Evaluate the logistic cumulative probability
Stata reference · R reference 
display logisticden(1,2,3)
 
stats::dlogis(3, location=1, scale=2)
stats 		Evaluate the logistic density
Stata reference · R reference 
display logistictail(1,2,3)
 
stats::plogis(3, location=1, scale=2, lower.tail=FALSE)
stats 		Evaluate the logistic upper-tail probability
Stata reference · R reference 
display nF(4,20,2,1.5)
 
stats::pf(1.5, df1=4, df2=20, ncp=2)
stats 		Evaluate the nF noncentral CDF
Stata reference · R reference 
display nFden(4,20,2,1.5)
 
stats::df(1.5, df1=4, df2=20, ncp=2)
stats 		Evaluate the nF noncentral density
Stata reference · R reference 
display nFtail(4,20,2,1.5)
 
stats::pf(1.5, df1=4, df2=20, ncp=2, lower.tail=FALSE)
stats 		Evaluate the nF noncentral upper tail
Stata reference · R reference 
display nbetaden(2,3,2,0.4)
 
stats::dbeta(0.4, shape1=2, shape2=3, ncp=2)
stats 		Evaluate the nibeta noncentral density
Stata reference · R reference 
display nbinomial(4,3,0.6)
 
stats::pnbinom(3, size=4, prob=0.6)
stats 		Compute an inclusive nbinomial lower tail
Stata reference · R reference 
display nbinomialp(4,3,0.6)
 
stats::dnbinom(3, size=4, prob=0.6)
stats 		Evaluate a nbinomial probability mass
Stata reference · R reference 
display nbinomialtail(4,3,0.6)
 
stats::pnbinom(2, size=4, prob=0.6, lower.tail=FALSE)
stats 		Compute an inclusive nbinomial upper tail
Stata includes k in the upper tail; R lower.tail=FALSE excludes its q argument, so use k-1.
Stata reference · R reference 
display nchi2(4,2,3)
 
stats::pchisq(3, df=4, ncp=2)
stats 		Evaluate the nchi2 noncentral CDF
Stata reference · R reference 
display nchi2den(4,2,3)
 
stats::dchisq(3, df=4, ncp=2)
stats 		Evaluate the nchi2 noncentral density
Stata reference · R reference 
display nchi2tail(4,2,3)
 
stats::pchisq(3, df=4, ncp=2, lower.tail=FALSE)
stats 		Evaluate the nchi2 noncentral upper tail
Stata reference · R reference 
display nibeta(2,3,2,0.4)
 
stats::pbeta(0.4, shape1=2, shape2=3, ncp=2)
stats 		Evaluate the nibeta noncentral CDF
Stata reference · R reference 
display normal(0.7)
 
stats::pnorm(0.7)
stats 		Evaluate the standard normal cumulative probability
Stata reference · R reference 
display normalden(0.7)
 
stats::dnorm(0.7)
stats 		Evaluate the standard normal density
Stata reference · R reference 
display normalden(3,1,2)
 
stats::dnorm(3, mean=1, sd=2)
stats 		Evaluate a normal density with location and scale
Stata reference · R reference 
display npnF(4,20,2,0.5)
 
stats::uniroot(function(ncp) stats::pf(2,df1=4,df2=20,ncp=ncp)-0.5,c(0,50),tol=1e-10)$root
stats 		Solve a noncentral F probability for noncentrality
Differs from Stata: Numerically inverts the noncentral F CDF over the displayed bracket. Other inputs can require a different bracket; solver tolerances differ from Stata.
Stata reference · R reference 
display npnchi2(4,8,0.5)
 
stats::uniroot(function(ncp) stats::pchisq(8,df=4,ncp=ncp)-0.5,c(0,50),tol=1e-10)$root
stats 		Solve a noncentral chi-squared probability for noncentrality
Differs from Stata: Numerically inverts the noncentral chi-square CDF over the displayed noncentrality bracket. Other inputs may need a wider bracket; Stata performs its own domain validation and inversion.
Stata reference · R reference 
display npnt(20,2,0.5)
 
stats::uniroot(function(ncp) stats::pt(2,df=20,ncp=ncp)-0.5,c(0,4),tol=1e-10)$root
stats 		Solve a noncentral t probability for noncentrality
Differs from Stata: Numerically inverts the noncentral t CDF over the displayed bracket. Negative or larger noncentralities require different bounds; solver tolerances differ from Stata.
Stata reference · R reference 
display nt(20,2,1.5)
 
stats::pt(1.5, df=20, ncp=2)
stats 		Evaluate the nt noncentral CDF
Stata reference · R reference 
display nttail(20,2,1.5)
 
stats::pt(1.5, df=20, ncp=2, lower.tail=FALSE)
stats 		Evaluate the nt noncentral upper tail
Stata reference · R reference 
display poisson(2,3)
 
stats::ppois(3, lambda=2)
stats 		Compute an inclusive poisson lower tail
Stata reference · R reference 
display poissonp(2,3)
 
stats::dpois(3, lambda=2)
stats 		Evaluate a poisson probability mass
Stata reference · R reference 
display poissontail(2,3)
 
stats::ppois(2, lambda=2, lower.tail=FALSE)
stats 		Compute an inclusive poisson upper tail
Stata includes k in the upper tail; R lower.tail=FALSE excludes its q argument, so use k-1.
Stata reference · R reference 
generate draw = rbeta(2,3)
 
stats::rbeta(1000,2,3)
stats 		Simulate rbeta variates
Differs from Stata: The same seed does not produce matching draws across Stata and R.
Stata reference · R reference 
generate draw = rbinomial(10,0.4)
 
stats::rbinom(1000,10,0.4)
stats 		Simulate rbinomial variates
Differs from Stata: The same seed does not produce matching draws across Stata and R.
Stata reference · R reference 
generate draw = rcauchy(1,2)
 
stats::rcauchy(1000,1,2)
stats 		Simulate rcauchy variates
Differs from Stata: The same seed does not produce matching draws across Stata and R.
Stata reference · R reference 
generate draw = rchi2(4)
 
stats::rchisq(1000,4)
stats 		Simulate rchi2 variates
Differs from Stata: The same seed does not produce matching draws across Stata and R.
Stata reference · R reference 
generate draw = rexponential(2)
 
stats::rexp(1000,rate=1/2)
stats 		Simulate rexponential variates
Differs from Stata: The same seed does not produce matching draws across Stata and R.
Stata reference · R reference 
generate draw = rgamma(3,2)
 
stats::rgamma(1000,shape=3,scale=2)
stats 		Simulate rgamma variates
Differs from Stata: The same seed does not produce matching draws across Stata and R.
Stata reference · R reference 
generate draw = rhypergeometric(20,8,5)
 
stats::rhyper(1000,m=8,n=12,k=5)
stats 		Simulate rhypergeometric variates
Differs from Stata: The same seed does not produce matching draws across Stata and R.
Stata reference · R reference 
generate draw = rlogistic(1,2)
 
stats::rlogis(1000,location=1,scale=2)
stats 		Simulate rlogistic variates
Differs from Stata: The same seed does not produce matching draws across Stata and R.
Stata reference · R reference 
generate draw = rnbinomial(4,0.6)
 
stats::rnbinom(1000,size=4,prob=0.6)
stats 		Simulate rnbinomial variates
Differs from Stata: The same seed does not produce matching draws across Stata and R.
Stata reference · R reference 
generate draw = rpoisson(2)
 
stats::rpois(1000,lambda=2)
stats 		Simulate rpoisson variates
Differs from Stata: The same seed does not produce matching draws across Stata and R.
Stata reference · R reference 
generate draw = rt(20)
 
stats::rt(1000,df=20)
stats 		Simulate rt variates
Differs from Stata: The same seed does not produce matching draws across Stata and R.
Stata reference · R reference 
generate draw = rweibull(2,3)
 
stats::rweibull(1000,shape=2,scale=3)
stats 		Simulate rweibull variates
Differs from Stata: The same seed does not produce matching draws across Stata and R.
Stata reference · R reference 
display t(10,1.2)
 
stats::pt(1.2, df=10)
stats 		Evaluate the Student t cumulative probability
Stata reference · R reference 
display tden(10,1.2)
 
stats::dt(1.2, df=10)
stats 		Evaluate the Student t density
Stata reference · R reference 
display ttail(10,1.2)
 
stats::pt(1.2, df=10, lower.tail=FALSE)
stats 		Evaluate the Student t upper-tail probability
Stata reference · R reference 
display tukeyprob(4,20,3)
 
stats::ptukey(3, nmeans=4, df=20)
stats 		Evaluate a studentized-range cumulative probability
Stata reference · R reference 
display weibull(2,3,1,5)
 
stats::pweibull(5-1, shape=2, scale=3)
stats 		Evaluate a shifted Weibull cumulative probability
Stata reference · R reference 
display weibull(2,3,4)
 
stats::pweibull(4, shape=2, scale=3)
stats 		Evaluate the Weibull cumulative probability
Stata reference · R reference 
display weibullden(2,3,4)
 
stats::dweibull(4, shape=2, scale=3)
stats 		Evaluate the Weibull density
Stata reference · R reference 
display weibullph(2,0.5,3)
 
stats::pweibull(3, shape=2, scale=0.5^(-1/2))
stats 		Use the proportional-hazards Weibull parameterization
Stata reference · R reference 
display weibullphden(2,0.5,3)
 
stats::dweibull(3, shape=2, scale=0.5^(-1/2))
stats 		Evaluate a PH Weibull density
Stata reference · R reference 
display weibullphtail(2,0.5,3)
 
stats::pweibull(3, shape=2, scale=0.5^(-1/2), lower.tail=FALSE)
stats 		Evaluate a PH Weibull survivor probability
Stata reference · R reference 
display weibulltail(2,3,4)
 
stats::pweibull(4, shape=2, scale=3, lower.tail=FALSE)
stats 		Evaluate the Weibull upper-tail probability
Stata reference · R reference

Models and estimation

Stata R Purpose, qualifications and sources 
betareg share x, link(cloglog)
 
betareg::betareg(share~x,data=data,link="cloglog",type="ML",link.phi="log")
betareg 		Fit beta regression with a cloglog mean link
Differs from Stata: Uses ordinary ML with a log link for precision, matching Stata slink(log). Without link.phi="log", betareg reports constant phi on the identity scale while Stata reports ln(phi). Responses must lie strictly inside (0,1).
Stata reference · R reference 
betareg share x, link(logit)
 
betareg::betareg(share~x,data=data,link="logit",type="ML",link.phi="log")
betareg 		Fit beta regression with a logit mean link
Differs from Stata: Uses ordinary ML with a log link for precision, matching Stata slink(log). Without link.phi="log", betareg reports constant phi on the identity scale while Stata reports ln(phi). Responses must lie strictly inside (0,1).
Stata reference · R reference 
betareg share x, link(probit)
 
betareg::betareg(share~x,data=data,link="probit",type="ML",link.phi="log")
betareg 		Fit beta regression with a probit mean link
Differs from Stata: Uses ordinary ML with a log link for precision, matching Stata slink(log). Without link.phi="log", betareg reports constant phi on the identity scale while Stata reports ln(phi). Responses must lie strictly inside (0,1).
Stata reference · R reference 
betareg share x, scale(z) slink(log)
 
betareg::betareg(share~x|z,data=data,link="logit",link.phi="log",type="ML")
betareg 		Model beta-regression precision using a covariate
Differs from Stata: Uses a log link for precision; align Stata scale-link parameterization before numerical comparison.
Stata reference · R reference 
binreg b i.g, or
 
stats::glm(b ~ factor(g),data=data,family=stats::binomial(link="logit"))
stats 		Fit binomial regression on the or scale
Differs from Stata: Risk-ratio and risk-difference models can encounter boundary and convergence problems; inspect fitted probabilities.
Stata reference · R reference 
binreg b i.g, rd
 
stats::glm(b ~ factor(g),data=data,family=stats::binomial(link="identity"))
stats 		Fit binomial regression on the rd scale
Differs from Stata: Risk-ratio and risk-difference models can encounter boundary and convergence problems; inspect fitted probabilities.
Stata reference · R reference 
binreg b i.g, rr
 
stats::glm(b ~ factor(g),data=data,family=stats::binomial(link="log"))
stats 		Fit binomial regression on the rr scale
Differs from Stata: Risk-ratio and risk-difference models can encounter boundary and convergence problems; inspect fitted probabilities.
Stata reference · R reference 
boxcox y x z, model(lhsonly)
 
MASS::boxcox(y ~ x+z,data=data,lambda=seq(-1,1,by=0.1),plotit=FALSE)
MASS 		Profile the Box-Cox response transformation parameter
Profiles a grid of response-transform parameters; this does not implement all Stata boxcox predictor-transform models.
Stata reference · R reference 
bsqreg y x z, reps(99)
 
{ fit <- quantreg::rq(y~x+z,data=data,tau=0.5); summary(fit,se="boot",R=99) }
quantreg 		Bootstrap quantile-regression inference
Differs from Stata: Uses 99 bootstrap replications for quantile-regression inference. R and Stata random streams and resampling implementations differ, so bootstrap standard errors will not be identical.
Stata reference · R reference 
cloglog b x z
 
stats::glm(b ~ x + z,data=data,family=stats::binomial(link="cloglog"))
stats 		Fit cloglog regression
Differs from Stata: R uses the complementary log-log link. Match the event category and complete cases; the GLM result object and convergence tolerances differ from Stata cloglog.
Stata reference · R reference 
cloglog b x z, noconstant
 
stats::glm(b ~ x + z - 1,data=data,family=stats::binomial(link="cloglog"))
stats 		Fit cloglog without an intercept
Differs from Stata: R uses the complementary log-log link. Match the event category and complete cases; the GLM result object and convergence tolerances differ from Stata cloglog.
Stata reference · R reference 
constraint define 1 x = z
cnsreg y x z, constraints(1)
 
{ fit <- stats::lm(y~I(x+z),data=data); c(intercept=unname(stats::coef(fit)[1]),x=unname(stats::coef(fit)[2]),z=unname(stats::coef(fit)[2])) }
stats 		Fit OLS with two coefficients constrained to equality
Fixed workflow. First define constraint 1 x = z in Stata. Reparameterization imposes this fixed linear constraint; covariance reporting needs the corresponding coefficient transformation.
Stata reference · R reference 
constraint define 1 x = z
 
L <- matrix(c(0,1,-1),nrow=1,dimnames=list(NULL,c("intercept","x","z")))
stats 		Represent a linear equality constraint as a matrix
Fixed workflow: L beta = 0 encodes equal x and z coefficients. A fitting routine must enforce the constraint.
Stata reference · R reference 
eivreg y x, reliability(x 0.8)
 
{ slope <- stats::cov(data$x,data$y)/(0.8*stats::var(data$x)); c(intercept=mean(data$y)-slope*mean(data$x),x=slope) }
stats 		Correct a simple regression slope for known predictor reliability
Coefficient-only workflow for one error-prone predictor with known reliability and classical independent measurement error; no EIV standard errors are supplied.
Stata reference · R reference 
fracreg logit share x
 
{ fit <- stats::glm(share ~ x,data=data,family=stats::quasibinomial(link="logit")); lmtest::coeftest(fit,vcov.=sandwich::vcovHC(fit,type="HC0"),df=Inf) }
sandwich 		Fit fractional logit regression with sandwich inference
Differs from Stata: Quasi-binomial fit avoids a noninteger-success warning. Sandwich inference is explicit; finite-sample corrections can differ.
Stata reference · R reference 
fracreg probit share x
 
{ fit <- stats::glm(share ~ x,data=data,family=stats::quasibinomial(link="probit")); lmtest::coeftest(fit,vcov.=sandwich::vcovHC(fit,type="HC0"),df=Inf) }
sandwich 		Fit fractional probit regression with sandwich inference
Differs from Stata: Quasi-binomial fit avoids a noninteger-success warning. Sandwich inference is explicit; finite-sample corrections can differ.
Stata reference · R reference 
glm b x z, family(binomial) link(cloglog)
 
stats::glm(b ~ x + z, data=data, family=stats::binomial(link="cloglog"))
stats 		Fit a binomial complementary-log-log GLM
Differs from Stata: The family and link are explicit. R and Stata can use different dispersion scaling and reference tests even when fitted means agree; align the sample and scale option before comparing standard errors.
Stata reference · R reference 
glm b x z, family(binomial) link(logit)
 
stats::glm(b ~ x + z, data=data, family=stats::binomial())
stats 		Fit a binomial logit-link GLM
Differs from Stata: The family and link are explicit. R and Stata can use different dispersion scaling and reference tests even when fitted means agree; align the sample and scale option before comparing standard errors.
Stata reference · R reference 
glm b x z, family(binomial) link(probit)
 
stats::glm(b ~ x + z, data=data, family=stats::binomial(link="probit"))
stats 		Fit a binomial probit-link GLM
Differs from Stata: The family and link are explicit. R and Stata can use different dispersion scaling and reference tests even when fitted means agree; align the sample and scale option before comparing standard errors.
Stata reference · R reference 
glm count x z, family(poisson) link(log)
 
stats::glm(count ~ x + z, data=data, family=stats::poisson())
stats 		Fit a Poisson log-link GLM
Differs from Stata: The family and link are explicit. R and Stata can use different dispersion scaling and reference tests even when fitted means agree; align the sample and scale option before comparing standard errors.
Stata reference · R reference 
glm count x, family(poisson) link(identity)
 
stats::glm(count ~ x, data=data, family=stats::poisson(link="identity"))
stats 		Fit a Poisson identity-link GLM
Differs from Stata: The family and link are explicit. R and Stata can use different dispersion scaling and reference tests even when fitted means agree; align the sample and scale option before comparing standard errors.
Stata reference · R reference 
glm events x, family(binomial trials) link(logit)
 
stats::glm(cbind(events,trials-events) ~ x,data=data,family=stats::binomial())
stats 		Fit grouped-binomial events and trials
Differs from Stata: The family and link are explicit. R and Stata can use different dispersion scaling and reference tests even when fitted means agree; align the sample and scale option before comparing standard errors.
Stata reference · R reference 
glm y x z, family(gamma) link(log)
 
stats::glm(y ~ x + z, data=data, family=stats::Gamma(link="log"))
stats 		Fit a gamma log-link model for positive outcomes
Differs from Stata: The family and link are explicit. R and Stata can use different dispersion scaling and reference tests even when fitted means agree; align the sample and scale option before comparing standard errors.
Stata reference · R reference 
glm y x z, family(gaussian) link(identity)
 
stats::glm(y ~ x + z, data=data, family=stats::gaussian())
stats 		Fit a Gaussian identity-link GLM
Differs from Stata: The family and link are explicit. R and Stata can use different dispersion scaling and reference tests even when fitted means agree; align the sample and scale option before comparing standard errors.
Stata reference · R reference 
glm y x z, family(gaussian) link(log)
 
stats::glm(y ~ x + z, data=data, family=stats::gaussian(link="log"))
stats 		Fit a Gaussian log-link GLM
Differs from Stata: The family and link are explicit. R and Stata can use different dispersion scaling and reference tests even when fitted means agree; align the sample and scale option before comparing standard errors.
Stata reference · R reference 
glm y x, family(gamma) link(identity)
 
stats::glm(y ~ x, data=data, family=stats::Gamma(link="identity"))
stats 		Fit a gamma identity-link model
Differs from Stata: The family and link are explicit. R and Stata can use different dispersion scaling and reference tests even when fitted means agree; align the sample and scale option before comparing standard errors.
Stata reference · R reference 
glm y x, family(gamma) link(inverse)
 
stats::glm(y ~ x, data=data, family=stats::Gamma(link="inverse"))
stats 		Fit a gamma inverse-link model
Differs from Stata: The family and link are explicit. R and Stata can use different dispersion scaling and reference tests even when fitted means agree; align the sample and scale option before comparing standard errors.
Stata reference · R reference 
glm y x, family(igaussian) link(identity)
 
stats::glm(y ~ x, data=data, family=stats::inverse.gaussian(link="identity"))
stats 		Fit an inverse Gaussian identity-link model
Differs from Stata: The family and link are explicit. R and Stata can use different dispersion scaling and reference tests even when fitted means agree; align the sample and scale option before comparing standard errors.
Stata reference · R reference 
glm y x, family(igaussian) link(log)
 
stats::glm(y ~ x, data=data, family=stats::inverse.gaussian(link="log"))
stats 		Fit an inverse Gaussian log-link model
Differs from Stata: The family and link are explicit. R and Stata can use different dispersion scaling and reference tests even when fitted means agree; align the sample and scale option before comparing standard errors.
Stata reference · R reference 
hetregress y x, het(z)
 
nlme::gls(y~x,data=data,weights=nlme::varExp(form=~z),method="ML")
nlme 		Fit a normal mean model with multiplicative heteroskedasticity
Differs from Stata: nlme models SD as exp(theta*z), so the coefficient on z in Stata log variance is 2*theta.
Stata reference · R reference 
intreg lower upper x z
 
survival::survreg(survival::Surv(lower,upper,type="interval2")~x+z,data=data,dist="gaussian")
survival 		Fit normal regression to interval-valued outcomes
Differs from Stata: Uses survreg as a Gaussian interval-likelihood engine for a cross-sectional outcome, without a time-to-event interpretation.
Stata reference · R reference 
iqreg y x z, quantiles(25 75)
 
{ fit <- quantreg::rq(y ~ x+z,data=data,tau=c(0.25,0.75)); stats::coef(fit)[,2]-stats::coef(fit)[,1] }
quantreg 		Estimate slopes of the interquartile range
Coefficient workflow only; obtain standard errors by jointly resampling both quantile fits.
Stata reference · R reference 
logistic b x z
 
{ fit <- stats::glm(b ~ x+z,data=data,family=stats::binomial()); exp(stats::coef(fit)) }
stats 		Report logistic odds ratios
Differs from Stata: Exponentiates the binomial logit coefficients to odds ratios. Returns point estimates only, without the confidence intervals and tests printed by Stata logistic.
Stata reference · R reference 
logit b x z
 
stats::glm(b ~ x + z,data=data,family=stats::binomial(link="logit"))
stats 		Fit logit regression
Differs from Stata: R uses a binomial GLM with the logit link and returns coefficients on the log-odds scale. Match binary coding and complete cases; Stata iteration diagnostics and stored result structure differ.
Stata reference · R reference 
logit b x z, noconstant
 
stats::glm(b ~ x + z - 1,data=data,family=stats::binomial(link="logit"))
stats 		Fit logit without an intercept
Differs from Stata: R uses a binomial GLM with the logit link and returns coefficients on the log-odds scale. Match binary coding and complete cases; Stata iteration diagnostics and stored result structure differ.
Stata reference · R reference 
mlexp (lnnormalden(y,{mu},exp({lnsigma})))
 
stats::optim(c(mu=mean(data$y),log_sigma=log(stats::sd(data$y))),function(par) -sum(stats::dnorm(data$y,mean=par[1],sd=exp(par[2]),log=TRUE)),method="BFGS",hessian=TRUE)
stats 		Maximize an explicitly defined normal likelihood
Differs from Stata: R optimizes log sigma to enforce positivity; translate the scale parameter back using exp().
Stata reference · R reference 
mlogit g x z, baseoutcome(4)
 
nnet::multinom(stats::relevel(factor(g), ref = "4") ~ x+z,data=data,trace=FALSE)
nnet 		Fit multinomial logit with a chosen reference outcome
Differs from Stata: Relevels outcome category 4 to the first R factor level, matching baseoutcome(4). Coefficient labels and optimization criteria differ; keep the same observations and category support.
Stata reference · R reference 
nbreg count x z
 
MASS::glm.nb(count ~ x+z,data=data)
MASS 		Fit NB2 negative-binomial regression
Differs from Stata: Both use NB2 variance mu+alpha*mu^2; MASS theta is 1/alpha.
Stata reference · R reference 
nl (y = {a=1} + {b=2}*x/({c=1}+x))
 
stats::nls(y ~ a+b*x/(c+x),data=data,start=list(a=1,b=2,c=1))
stats 		Fit a nonlinear saturating response curve
Differs from Stata: Fits this explicit nonlinear mean function from the supplied starting values. nls and Stata nl use different convergence controls and can reach different local solutions.
Stata reference · R reference 
nl (y = {a=2}*exp({b=0.2}*x))
 
stats::nls(y ~ a*exp(b*x),data=data,start=list(a=2,b=0.2))
stats 		Fit a nonlinear exponential mean curve
Differs from Stata: Fits this explicit nonlinear mean function from the supplied starting values. nls and Stata nl use different convergence controls and can reach different local solutions.
Stata reference · R reference 
ologit y x z
 
MASS::polr(y ~ x+z,data=data,method="logistic",Hess=TRUE)
MASS 		Fit ordered ologit regression
Differs from Stata: MASS uses cumulative link intercepts; compare cutpoint and slope sign conventions.
Stata reference · R reference 
oprobit y x z
 
MASS::polr(y ~ x+z,data=data,method="probit",Hess=TRUE)
MASS 		Fit ordered oprobit regression
Differs from Stata: MASS uses cumulative link intercepts; compare cutpoint and slope sign conventions.
Stata reference · R reference 
poisson count x z, exposure(exposure)
 
stats::glm(count ~ x+z+offset(log(exposure)),data=data,family=stats::poisson())
stats 		Fit Poisson rates with exposure
Differs from Stata: Uses the explicit log offset where shown; exposure must be strictly positive. R returns a glm object or exponentiated coefficients, rather than the Stata estimation table; align overdispersion and robust variance options.
Stata reference · R reference 
poisson count x z, irr
 
{ fit <- stats::glm(count~x+z,data=data,family=stats::poisson()); exp(stats::coef(fit)) }
stats 		Report Poisson incidence-rate ratios
Differs from Stata: Uses the explicit log offset where shown; exposure must be strictly positive. R returns a glm object or exponentiated coefficients, rather than the Stata estimation table; align overdispersion and robust variance options.
Stata reference · R reference 
poisson count x z, noconstant
 
stats::glm(count ~ x+z-1,data=data,family=stats::poisson())
stats 		Fit Poisson regression without an intercept
Differs from Stata: Uses the explicit log offset where shown; exposure must be strictly positive. R returns a glm object or exponentiated coefficients, rather than the Stata estimation table; align overdispersion and robust variance options.
Stata reference · R reference 
probit b x z
 
stats::glm(b ~ x + z,data=data,family=stats::binomial(link="probit"))
stats 		Fit probit regression
Differs from Stata: R uses the probit binomial GLM. Match binary coding and complete cases; the GLM result object and convergence tolerances differ from Stata probit.
Stata reference · R reference 
probit b x z, noconstant
 
stats::glm(b ~ x + z - 1,data=data,family=stats::binomial(link="probit"))
stats 		Fit probit without an intercept
Differs from Stata: R uses the probit binomial GLM. Match binary coding and complete cases; the GLM result object and convergence tolerances differ from Stata probit.
Stata reference · R reference 
qreg y x z, quantile(25)
 
quantreg::rq(y ~ x+z,data=data,tau=0.25)
quantreg 		Fit conditional quantile 25 regression
Differs from Stata: Fits the selected conditional quantile. rq and Stata qreg use different optimization and default covariance estimators; coefficient agreement does not establish standard-error parity.
Stata reference · R reference 
qreg y x z, quantile(75)
 
quantreg::rq(y ~ x+z,data=data,tau=0.75)
quantreg 		Fit conditional quantile 75 regression
Differs from Stata: Fits the selected conditional quantile. rq and Stata qreg use different optimization and default covariance estimators; coefficient agreement does not establish standard-error parity.
Stata reference · R reference 
regress y c.x##c.x
 
stats::lm(y ~ x + I(x^2), data=data)
stats 		Fit a quadratic regression
Differs from Stata: OLS coefficient correspondence requires the same complete-case sample, contrasts and weights. R returns an lm object; Stata coefficient tables and stored estimates have a different structure.
Stata reference · R reference 
regress y c.x##c.z
 
stats::lm(y ~ x * z, data=data)
stats 		Include two continuous main effects and their interaction
Differs from Stata: OLS coefficient correspondence requires the same complete-case sample, contrasts and weights. R returns an lm object; Stata coefficient tables and stored estimates have a different structure.
Stata reference · R reference 
regress y c.x##i.g
 
stats::lm(y ~ x * factor(g), data=data)
stats 		Fit separate linear slopes by group
Differs from Stata: OLS coefficient correspondence requires the same complete-case sample, contrasts and weights. R returns an lm object; Stata coefficient tables and stored estimates have a different structure.
Stata reference · R reference 
regress y i.g c.x
 
stats::lm(y ~ factor(g) + x, data=data)
stats 		Adjust regression for a categorical predictor
Differs from Stata: OLS coefficient correspondence requires the same complete-case sample, contrasts and weights. R returns an lm object; Stata coefficient tables and stored estimates have a different structure.
Stata reference · R reference 
regress y x [aweight=w]
 
stats::lm(y ~ x, data=data, weights=w)
stats 		Fit analytic-weighted least squares
Differs from Stata: Weights are proportional inverse variances. Check scale and residual degrees of freedom before comparing software.
Stata reference · R reference 
regress y x z if b == 1
 
stats::lm(y ~ x + z, data=data, subset=b == 1)
stats 		Fit linear regression in a specified subgroup
Differs from Stata: OLS coefficient correspondence requires the same complete-case sample, contrasts and weights. R returns an lm object; Stata coefficient tables and stored estimates have a different structure.
Stata reference · R reference 
regress y x z, beta
 
{ fit <- stats::lm(y ~ x + z, data=data); stats::coef(fit)[c("x","z")] * vapply(data[c("x","z")], stats::sd, numeric(1)) / stats::sd(data$y) }
stats 		Report standardized linear regression slopes
Differs from Stata: OLS coefficient correspondence requires the same complete-case sample, contrasts and weights. R returns an lm object; Stata coefficient tables and stored estimates have a different structure.
Stata reference · R reference 
regress y x z, noconstant
 
stats::lm(y ~ x + z - 1, data=data)
stats 		Fit linear regression through the origin
Differs from Stata: OLS coefficient correspondence requires the same complete-case sample, contrasts and weights. R returns an lm object; Stata coefficient tables and stored estimates have a different structure.
Stata reference · R reference 
regress y x z, vce(cluster g)
 
{ fit <- stats::lm(y ~ x + z, data = data, subset = !is.na(g)); lmtest::coeftest(fit, vcov. = sandwich::vcovCL(fit, cluster = ~g, type = "HC1"), df = length(unique(stats::expand.model.frame(fit, ~g)$g)) - 1L) }
sandwich 		Use one-way clustered standard errors with cluster degrees of freedom
Differs from Stata: Only three clusters in this demonstration fixture; substantive applications need adequate independent clusters.
Stata reference · R reference 
regress y x z, vce(hc2)
 
{ fit <- stats::lm(y~x+z,data=data); lmtest::coeftest(fit, vcov.=sandwich::vcovHC(fit,type="HC2")) }
sandwich 		Report HC2 heteroskedasticity-robust OLS inference
Differs from Stata: Coefficient estimates are OLS; covariance correction is selected explicitly.
Stata reference · R reference 
regress y x z, vce(hc3)
 
{ fit <- stats::lm(y~x+z,data=data); lmtest::coeftest(fit, vcov.=sandwich::vcovHC(fit,type="HC3")) }
sandwich 		Report HC3 heteroskedasticity-robust OLS inference
Differs from Stata: Coefficient estimates are OLS; covariance correction is selected explicitly.
Stata reference · R reference 
regress y x z, vce(robust)
 
{ fit <- stats::lm(y~x+z,data=data); lmtest::coeftest(fit, vcov.=sandwich::vcovHC(fit,type="HC1")) }
sandwich 		Report HC1 heteroskedasticity-robust OLS inference
Differs from Stata: Coefficient estimates are OLS; covariance correction is selected explicitly.
Stata reference · R reference 
rreg y x z
 
MASS::rlm(y ~ x+z,data=data,psi=MASS::psi.bisquare,init="lts",maxit=100)
MASS 		Fit a resistant bisquare regression
Workflow equivalent: Stata rreg uses its own screening and iteration sequence, so coefficients need not match.
Stata reference · R reference 
sqreg y x z, quantiles(25 50 75)
 
quantreg::rq(y ~ x+z,data=data,tau=c(0.25,0.5,0.75))
quantreg 		Fit several conditional quantiles together
This fits coefficients; Stata sqreg also bootstraps their joint covariance. Use a shared-resample bootstrap for joint inference.
Stata reference · R reference 
tobit y x z, ll(0)
 
AER::tobit(y~x+z,data=data,left=0)
AER 		Fit a left-censored normal Tobit model
Differs from Stata: Uses the explicit left and/or right censoring limits. AER returns a survreg-based Gaussian censored model; scale parameter reporting and coefficient-table inference differ from Stata.
Stata reference · R reference 
tobit y x z, ll(0) ul(2)
 
AER::tobit(y~x+z,data=data,left=0,right=2)
AER 		Fit a two-limit censored normal model
Differs from Stata: Uses the explicit left and/or right censoring limits. AER returns a survreg-based Gaussian censored model; scale parameter reporting and coefficient-table inference differ from Stata.
Stata reference · R reference 
tobit y x z, ul(2)
 
AER::tobit(y~x+z,data=data,left=-Inf,right=2)
AER 		Fit a right-censored normal Tobit model
Differs from Stata: Uses the explicit left and/or right censoring limits. AER returns a survreg-based Gaussian censored model; scale parameter reporting and coefficient-table inference differ from Stata.
Stata reference · R reference 
zinb count x, inflate(z)
 
pscl::zeroinfl(count~x|z,data=data,dist="negbin",link="logit")
pscl 		Fit a zero-inflated NB2 count model
Differs from Stata: Negative-binomial theta is reciprocal to Stata alpha in the NB2 component.
Stata reference · R reference 
zip count x, inflate(z)
 
pscl::zeroinfl(count~x|z,data=data,dist="poisson",link="logit")
pscl 		Fit zero-inflated Poisson with logit zero inflation
Differs from Stata: Fits the stated zero-inflated Poisson model with separate count and inflation equations. Match the inflation link, offsets and complete cases; optimizers and coefficient labels differ.
Stata reference · R reference 
zip count x, inflate(z) probit
 
pscl::zeroinfl(count~x|z,data=data,dist="poisson",link="probit")
pscl 		Fit zero-inflated Poisson with probit zero inflation
Differs from Stata: Fits the stated zero-inflated Poisson model with separate count and inflation equations. Match the inflation link, offsets and complete cases; optimizers and coefficient labels differ.
Stata reference · R reference

Postestimation and diagnostics

Stata R Purpose, qualifications and sources 
brier b p
 
mean((data$b-data$p)^2)
stats 		Calculate the overall Brier mean probability score
Differs from Stata: Returns the overall score only; Stata brier also reports decomposition components.
Stata reference · R reference 
dfbeta x
 
stats::dfbetas(fit)[,"x"]
stats 		Obtain standardized deletion influence on one slope
Differs from Stata: Returns standardized coefficient-deletion diagnostics from dfbetas for x; coefficient ordering, missing rows and diagnostic naming must be aligned with the Stata model.
Stata reference · R reference 
estat hettest x, iid
 
lmtest::bptest(fit,varformula=~x,data=data,studentize=TRUE)
lmtest 		Test heteroskedasticity against one predictor using a studentized statistic
Differs from Stata: Uses the studentized Koenker version of the Breusch-Pagan test, corresponding to the iid variant shown. Match the auxiliary regressors and estimation sample; the default bptest differs from default Stata hettest.
Stata reference · R reference 
estat hettest, rhs
 
lmtest::bptest(fit,studentize=FALSE)
lmtest 		Run the classical Breusch-Pagan heteroskedasticity test
Differs from Stata: studentize=FALSE selects the normal-theory version; default bptest uses the Koenker version.
Stata reference · R reference 
estat ic
 
c(AIC=stats::AIC(fit),BIC=stats::BIC(fit))
stats 		Compare penalized-likelihood model criteria
R keeps fitted model objects; model names are fixed workflow labels.
Stata reference · R reference 
estat imtest, white
 
lmtest::bptest(fit,varformula=~x+z+I(x^2)+I(z^2)+x:z,data=data,studentize=TRUE)
lmtest 		Run a White-style auxiliary-regression heteroskedasticity test
This covers the White heteroskedasticity component, not the complete information-matrix decomposition.
Stata reference · R reference 
estat ovtest
 
lmtest::resettest(fit,power=2:4,type="fitted")
lmtest 		Run Ramsey RESET using powers of fitted values
Differs from Stata: Uses RESET powers two through four of the fitted values or regressors as shown. Regressor-factor expansion and rank-deficient auxiliary terms may differ from Stata.
Stata reference · R reference 
estat ovtest, rhs
 
lmtest::resettest(fit,power=2:4,type="regressor")
lmtest 		Run RESET using powers of regressors
Differs from Stata: Uses RESET powers two through four of the fitted values or regressors as shown. Regressor-factor expansion and rank-deficient auxiliary terms may differ from Stata.
Stata reference · R reference 
estat summarize
 
summary(stats::model.frame(fit))
stats 		Summarize variables in the actual estimation sample
R keeps fitted model objects; model names are fixed workflow labels.
Stata reference · R reference 
estat vce
 
stats::vcov(fit)
stats 		Extract the coefficient covariance matrix
R keeps fitted model objects; model names are fixed workflow labels.
Stata reference · R reference 
estat vce, correlation
 
stats::cov2cor(stats::vcov(fit))
stats 		Extract correlations among coefficient estimates
R keeps fitted model objects; model names are fixed workflow labels.
Stata reference · R reference 
estat vif
 
car::vif(fit)
car 		Compute variance inflation factors after OLS
Differs from Stata: car returns ordinary VIFs for one-degree-of-freedom terms and generalized VIFs for multi-parameter terms. Stata estat vif reports column-level VIFs, so factor-term output is not directly interchangeable.
Stata reference · R reference 
estimates dir
 
names(models)
stats 		List stored model names
R keeps fitted model objects; model names are fixed workflow labels.
Stata reference · R reference 
estimates replay
 
summary(fit)
stats 		Redisplay a fitted linear model summary
R keeps fitted model objects; model names are fixed workflow labels.
Stata reference · R reference 
estimates restore model1
 
fit <- models[["model1"]]
stats 		Restore one model from a model collection
R keeps fitted model objects; model names are fixed workflow labels.
Stata reference · R reference 
estimates save model1
 
{ path <- tempfile(fileext=".rds"); saveRDS(fit,path); readRDS(path) }
base 		Save and reload a fitted model object
Writes to an R temporary file; choose an explicit project path for persistent storage.
Stata reference · R reference 
estimates stats model1 model2
 
do.call(stats::AIC,unname(models))
stats 		Compare AIC across stored model objects
R keeps fitted model objects; model names are fixed workflow labels.
Stata reference · R reference 
estimates store model1
 
model1 <- fit
stats 		Store a fitted model under a name
R keeps fitted model objects; model names are fixed workflow labels.
Stata reference · R reference 
estimates table model1 model2, b se
 
lapply(models,function(m) summary(m)$coefficients[,1:2,drop=FALSE])
stats 		Compare coefficients and standard errors across models
R keeps fitted model objects; model names are fixed workflow labels.
Stata reference · R reference 
lincom x + z
 
{ L <- c(0,1,1); c(estimate=sum(L*stats::coef(fit)),se=sqrt(drop(t(L)%*%stats::vcov(fit)%*%L))) }
stats 		Estimate the sum of two slopes with its standard error
Differs from Stata: The contrast vector assumes coefficient order intercept, x, z. Returns the estimate and its covariance-based standard error only; Stata also prints tests and confidence limits.
Stata reference · R reference 
lincom x - z
 
{ L <- c(0,1,-1); c(estimate=sum(L*stats::coef(fit)),se=sqrt(drop(t(L)%*%stats::vcov(fit)%*%L))) }
stats 		Estimate a difference of slopes with its standard error
Differs from Stata: The contrast vector assumes coefficient order intercept, x, z. Returns the estimate and its covariance-based standard error only; Stata also prints tests and confidence limits.
Stata reference · R reference 
linktest
 
{ data$hat <- stats::fitted(fit); stats::lm(y ~ hat+I(hat^2),data=data) }
stats 		Regress an outcome on fitted values and their square
Shown for OLS; link tests after other model families must refit the appropriate family.
Stata reference · R reference 
lrtest reduced full
 
lmtest::lrtest(reduced,full)
lmtest 		Compare nested maximum-likelihood models
Differs from Stata: Fixed workflow.
Stata reference · R reference 
margins, at(x=(2 3 4))
 
vapply(c(2,3,4),function(value) { newdata <- data; newdata$x <- value; mean(stats::predict(fit,newdata=newdata,type="response")) },numeric(1))
stats 		Predict at specified covariate values
Differs from Stata: At each x value, predictions are averaged over the observed distribution of z.
Stata reference · R reference 
margins, dydx(x)
 
{ X <- stats::model.matrix(fit); beta <- stats::coef(fit); p <- stats::fitted(fit); gradient <- beta["x"]*colMeans(X*as.vector(p*(1-p)*(1-2*p))); gradient["x"] <- gradient["x"]+mean(p*(1-p)); c(estimate=unname(beta["x"]*mean(p*(1-p))),se=sqrt(drop(t(gradient)%*%stats::vcov(fit)%*%gradient))) }
stats 		Average a continuous-predictor marginal effect
Differs from Stata: Uses the fitted GLM and its estimation data; reference grids and averaging rules should be chosen explicitly.
Stata reference · R reference 
margins, dydx(x) atmeans
 
{ X <- colMeans(stats::model.matrix(fit)); beta <- stats::coef(fit); p <- stats::plogis(sum(X*beta)); c(estimate=unname(beta["x"]*p*(1-p))) }
stats 		Evaluate a marginal effect at mean covariates
Differs from Stata: Returns the derivative at means; use a delta-method gradient if standard errors are needed.
Stata reference · R reference 
margins, expression(predict(xb))
 
{ L <- colMeans(stats::model.matrix(fit)); c(estimate=sum(L*stats::coef(fit)),se=sqrt(drop(t(L)%*%stats::vcov(fit)%*%L))) }
stats 		Average predictions on the linear-index scale
Differs from Stata: Uses the fitted GLM and its estimation data; reference grids and averaging rules should be chosen explicitly.
Stata reference · R reference 
nlcom _b[x]/_b[z]
 
car::deltaMethod(fit,"x / z")
car 		Use the delta method for a ratio of slopes
Differs from Stata: Uses the first-order delta method on named fitted coefficients. Returns the transformed estimate and standard error; Stata also prints reference tests and confidence limits.
Stata reference · R reference 
nlcom exp(_b[x])
 
car::deltaMethod(fit,"exp(x)")
car 		Use the delta method for an exponentiated slope
Differs from Stata: Uses the first-order delta method on named fitted coefficients. Returns the transformed estimate and standard error; Stata also prints reference tests and confidence limits.
Stata reference · R reference 
predict result, cooksd
 
data$result <- stats::cooks.distance(fit)
stats 		Obtain Cook influence distances
Differs from Stata: Run after fit <- lm(y ~ x + z, data=data).
Stata reference · R reference 
predict result, deviance
 
data$result <- stats::residuals(fit,type="deviance")
stats 		Compute binary-model deviance residuals
Differs from Stata: After fitting a binary-response GLM. Prediction options are model-specific.
Stata reference · R reference 
predict result, dfits
 
data$result <- stats::dffits(fit)
stats 		Obtain DFFITS influence statistics
Differs from Stata: Run after fit <- lm(y ~ x + z, data=data).
Stata reference · R reference 
predict result, hat
 
data$result <- stats::hatvalues(fit)
stats 		Obtain leverage values
Differs from Stata: Run after fit <- lm(y ~ x + z, data=data).
Stata reference · R reference 
predict result, pearson
 
data$result <- stats::residuals(fit,type="pearson")
stats 		Compute binary-model Pearson residuals
Differs from Stata: After fitting a binary-response GLM. Prediction options are model-specific.
Stata reference · R reference 
predict result, pr
 
data$result <- stats::predict(fit,type="response")
stats 		Predict logistic probabilities
Differs from Stata: After fitting a binary-response GLM. Prediction options are model-specific.
Stata reference · R reference 
predict result, residuals
 
data$result <- stats::residuals(fit)
stats 		Obtain raw OLS residuals
Differs from Stata: Run after fit <- lm(y ~ x + z, data=data).
Stata reference · R reference 
predict result, rstandard
 
data$result <- stats::rstandard(fit)
stats 		Obtain internally standardized residuals
Differs from Stata: Run after fit <- lm(y ~ x + z, data=data).
Stata reference · R reference 
predict result, rstudent
 
data$result <- stats::rstudent(fit)
stats 		Obtain externally studentized residuals
Differs from Stata: Run after fit <- lm(y ~ x + z, data=data).
Stata reference · R reference 
predict result, stdf
 
data$result <- sqrt(stats::predict(fit,se.fit=TRUE)$se.fit^2+summary(fit)$sigma^2)
stats 		Obtain forecast standard errors including residual variance
Differs from Stata: Adds residual variance to the squared standard error of the fitted OLS mean. This is a standard error for a new individual prediction, not the fitted mean; assumes homoskedastic errors and matching model sample.
Stata reference · R reference 
predict result, stdp
 
data$result <- stats::predict(fit,se.fit=TRUE)$se.fit
stats 		Obtain standard errors of fitted means
Differs from Stata: Run after fit <- lm(y ~ x + z, data=data).
Stata reference · R reference 
predict result, xb
 
data$result <- stats::predict(fit)
stats 		Obtain linear fitted values
Differs from Stata: Run after fit <- lm(y ~ x + z, data=data).
Stata reference · R reference 
predict result, xb
 
data$result <- stats::predict(fit,type="link")
stats 		Predict logistic linear indices
Differs from Stata: After fitting a binary-response GLM. Prediction options are model-specific.
Stata reference · R reference 
roccomp b x z
 
pROC::roc.test(pROC::roc(data$b,data$x,levels=c(0,1),direction="<",quiet=TRUE),pROC::roc(data$b,data$z,levels=c(0,1),direction="<",quiet=TRUE),paired=TRUE,method="delong")
pROC 		Compare paired marker ROC areas using DeLong inference
Differs from Stata: Both markers must be measured on the same complete-case subjects for paired inference.
Stata reference · R reference 
roctab b x
 
pROC::roc(data$b,data$x,levels=c(0,1),direction="<",quiet=TRUE)
pROC 		Estimate a nonparametric ROC area for a marker
Differs from Stata: Explicit direction means larger markers indicate cases; fixing direction prevents data-driven reversal.
Stata reference · R reference 
roctab b x, detail
 
pROC::coords(pROC::roc(data$b,data$x,levels=c(0,1),direction="<",quiet=TRUE),x="all",ret=c("threshold","sensitivity","specificity"))
pROC 		Tabulate sensitivity and specificity across ROC thresholds
Differs from Stata: Returns all pROC thresholds with sensitivity and specificity for explicit event direction. Threshold locations and endpoint conventions differ from Stata roctab detail cutpoints.
Stata reference · R reference 
test x = 0
 
car::linearHypothesis(fit,"x = 0")
car 		Test a single regression coefficient
Differs from Stata: Tests the specified coefficient restriction on the existing fit. car selects its reference statistic from the model class; a GLM chi-square test differs from the OLS F test used by Stata regress postestimation.
Stata reference · R reference 
test x = z
 
car::linearHypothesis(fit,"x = z")
car 		Test equality of two regression slopes
Differs from Stata: Tests the specified coefficient restriction on the existing fit. car selects its reference statistic from the model class; a GLM chi-square test differs from the OLS F test used by Stata regress postestimation.
Stata reference · R reference 
test x z
 
car::linearHypothesis(fit,c("x = 0","z = 0"))
car 		Jointly test two regression coefficients
Differs from Stata: Tests the specified coefficient restriction on the existing fit. car selects its reference statistic from the model class; a GLM chi-square test differs from the OLS F test used by Stata regress postestimation.
Stata reference · R reference 
testnl exp(_b[x]) = 1
 
{ d <- car::deltaMethod(fit,"exp(x)"); w <- ((d$Estimate-1)/d$SE)^2; c(chisq=w,p.value=stats::pchisq(w,df=1,lower.tail=FALSE)) }
car 		Test an exponentiated coefficient with a nonlinear Wald test
Differs from Stata: Forms a one-degree-of-freedom Wald chi-square from a first-order delta-method standard error. The R vector omits Stata coefficient and confidence-interval output.
Stata reference · R reference 
testparm i.g
 
car::linearHypothesis(fit,c("g6 = 0","g8 = 0"))
car 		Jointly test a categorical predictor
Differs from Stata: The restrictions name the fitted factor coefficients explicitly. Different reference levels or contrasts require different coefficient names; this is not an automatic expansion of any Stata factor-variable expression.
Stata reference · R reference

Hypothesis tests and confidence intervals

Stata R Purpose, qualifications and sources 
bitest b = 0.5
 
stats::binom.test(sum(data$b),nrow(data),p=0.5)
stats 		Test a binary success probability exactly
Differs from Stata: Returns an exact two-sided binomial test. The code assumes the displayed binary sample has no missing values; Stata additionally prints lower- and upper-tail probabilities.
Stata reference · R reference 
bitesti 20 7 0.5
 
stats::binom.test(7,20,p=0.5)
stats 		Run an exact binomial test from counts
Differs from Stata: Uses an exact two-sided binomial test from the supplied count and total. R reports one htest result; Stata additionally displays both one-sided probabilities.
Stata reference · R reference 
ci means y
 
stats::t.test(data$y)$conf.int
stats 		Compute a Student-t confidence interval for a mean
Differs from Stata: Returns only the selected confidence limits, not Stata sample-size and point-estimate columns. Match nonmissing sample size, confidence level and the explicit interval method.
Stata reference · R reference 
ci proportions b, exact
 
stats::binom.test(sum(data$b),nrow(data))$conf.int
stats 		Compute a Clopper-Pearson proportion interval
Differs from Stata: Returns only the selected confidence limits, not Stata sample-size and point-estimate columns. Match nonmissing sample size, confidence level and the explicit interval method.
Stata reference · R reference 
ci proportions b, wilson
 
stats::prop.test(sum(data$b),nrow(data),correct=FALSE)$conf.int
stats 		Compute a Wilson score interval for a proportion
Differs from Stata: Returns only the selected confidence limits, not Stata sample-size and point-estimate columns. Match nonmissing sample size, confidence level and the explicit interval method.
Stata reference · R reference 
ci variances y
 
(length(data$y)-1)*stats::var(data$y)/stats::qchisq(c(0.975,0.025),df=length(data$y)-1)
stats 		Compute a normal-theory variance interval
Differs from Stata: Returns only the selected confidence limits, not Stata sample-size and point-estimate columns. Match nonmissing sample size, confidence level and the explicit interval method.
Stata reference · R reference 
ci variances y, sd
 
sqrt((length(data$y)-1)*stats::var(data$y)/stats::qchisq(c(0.975,0.025),df=length(data$y)-1))
stats 		Compute a normal-theory standard-deviation interval
Differs from Stata: Returns only the selected confidence limits, not Stata sample-size and point-estimate columns. Match nonmissing sample size, confidence level and the explicit interval method.
Stata reference · R reference 
cii means 50 10 2
 
10+c(-1,1)*stats::qt(0.975,49)*2/sqrt(50)
stats 		Compute a mean interval from summary statistics
Differs from Stata: Computes confidence limits from the displayed summary inputs. The R result omits Stata sample-size and point-estimate columns; the stated method and confidence level determine the interval.
Stata reference · R reference 
cii proportions 20 7, exact
 
stats::binom.test(7,20)$conf.int
stats 		Compute an exact proportion interval from counts
Differs from Stata: Computes confidence limits from the displayed summary inputs. The R result omits Stata sample-size and point-estimate columns; the stated method and confidence level determine the interval.
Stata reference · R reference 
ksmirnov y, by(g)
 
stats::ks.test(data$y[data$g=="a"],data$y[data$g=="b"],exact=FALSE)
stats 		Compare two empirical distribution functions
Differs from Stata: The two-sample form uses a grouping variable; a one-sample test instead requires a fully specified reference CDF.
Stata reference · R reference 
ktau y x
 
stats::cor.test(data$y,data$x,method="kendall",exact=FALSE)
stats 		Test Kendall rank association
Differs from Stata: Uses the asymptotic Kendall test. Tie adjustments and the reported tau variant must match; R returns one htest object rather than Stata full concordance summaries.
Stata reference · R reference 
kwallis y, by(g)
 
stats::kruskal.test(y~g,data=data)
stats 		Test equality of distributions across groups
Differs from Stata: Computes the tie-corrected Kruskal-Wallis chi-square approximation. R returns an htest object without Stata group rank-sum summaries.
Stata reference · R reference 
median y, by(g)
 
stats::chisq.test(table(data$g,data$y>stats::median(data$y)),correct=FALSE)
stats 		Run a two-group median test
Differs from Stata: Tests independence of group and the above-pooled-median indicator. The example assigns observations tied at the median to the lower category; align Stata ties() handling before comparison.
Stata reference · R reference 
nptrend b, group(g) carmitage
 
stats::prop.trend.test(c(10,18,28),c(40,40,40),score=c(1,2,3))
stats 		Test an ordered trend in proportions
Differs from Stata: Illustrates aggregated successes and denominators at the three ordered group scores.
Stata reference · R reference 
oneway y g, bonferroni
 
stats::pairwise.t.test(data$y,data$g,p.adjust.method="bonferroni",pool.sd=TRUE)
stats 		Compare group means with Bonferroni adjustment
Differs from Stata: Uses the ordinary equal-variance one-way model or pooled pairwise t tests as shown. R object layout differs and does not reproduce every Stata tabulate column or comparison display.
Stata reference · R reference 
oneway y g, tabulate
 
stats::aov(y~factor(g),data=data)
stats 		Fit one-way analysis of variance
Differs from Stata: Uses the ordinary equal-variance one-way model or pooled pairwise t tests as shown. R object layout differs and does not reproduce every Stata tabulate column or comparison display.
Stata reference · R reference 
pcorr y x z
 
stats::cor(stats::residuals(stats::lm(y~z,data=data)),stats::residuals(stats::lm(x~z,data=data)))
stats 		Compute the partial correlation of y and x controlling z
Differs from Stata: This returns the y-x partial correlation; Stata pcorr also reports semipartial correlations for each listed predictor.
Stata reference · R reference 
prtest b == 0.5
 
stats::prop.test(sum(data$b),nrow(data),p=0.5,correct=FALSE)
stats 		Run an uncorrected large-sample one-proportion test
Differs from Stata: Disables the R continuity correction for the normal-approximation proportion test. R reports a chi-square statistic whose square root is the absolute Stata z; signed direction requires the group order.
Stata reference · R reference 
prtest b, by(g)
 
stats::prop.test(tapply(data$b,data$g,sum),as.vector(table(data$g)),correct=FALSE)
stats 		Compare two proportions without continuity correction
Differs from Stata: Disables the R continuity correction for the normal-approximation proportion test. R reports a chi-square statistic whose square root is the absolute Stata z; signed direction requires the group order.
Stata reference · R reference 
prtesti 100 0.4 120 0.5
 
stats::prop.test(c(40,60),c(100,120),correct=FALSE)
stats 		Compare proportions from aggregate sample sizes
Differs from Stata: Disables continuity correction. R returns a chi-square test; Stata reports a signed z and one-sided probabilities, although the two-sided p-values correspond for the same counts.
Stata reference · R reference 
ranksum y, by(g)
 
stats::wilcox.test(y~g,data=data,exact=FALSE,correct=FALSE)
stats 		Compare independent distributions with the rank-sum test
Differs from Stata: Aligns the requested exact or uncorrected asymptotic Wilcoxon calculation. Tie handling, group ordering and reported rank-sum versus U-statistic conventions must be aligned.
Stata reference · R reference 
ranksum y, by(g) exact
 
stats::wilcox.test(y~g,data=data,exact=TRUE)
stats 		Use exact rank-sum inference without ties
Differs from Stata: Aligns the requested exact or uncorrected asymptotic Wilcoxon calculation. Tie handling, group ordering and reported rank-sum versus U-statistic conventions must be aligned.
Stata reference · R reference 
sdtest y == 3
 
{ statistic <- (length(data$y)-1)*stats::var(data$y)/3^2; c(chisq=statistic,p.value=2*min(stats::pchisq(statistic,11),stats::pchisq(statistic,11,lower.tail=FALSE))) }
stats 		Test one normal population standard deviation
Differs from Stata: Uses a normal-theory variance-ratio or chi-square test. R returns test components in an htest object or vector; Stata also reports one-sided alternatives and additional confidence limits.
Stata reference · R reference 
sdtest y == x
 
stats::var.test(data$y,data$x)
stats 		Compare two unpaired sample variances
Differs from Stata: Uses a normal-theory variance-ratio or chi-square test. R returns test components in an htest object or vector; Stata also reports one-sided alternatives and additional confidence limits.
Stata reference · R reference 
sdtest y, by(g)
 
stats::var.test(y~g,data=data)
stats 		Compare variances of two independent normal samples
Differs from Stata: Uses a normal-theory variance-ratio or chi-square test. R returns test components in an htest object or vector; Stata also reports one-sided alternatives and additional confidence limits.
Stata reference · R reference 
sdtesti 30 2 40 3
 
{ f <- 2^2/3^2; c(F=f,p.value=2*min(stats::pf(f,29,39),stats::pf(f,29,39,lower.tail=FALSE))) }
stats 		Compare variances from sample sizes and standard deviations
Differs from Stata: Computes the two-sided normal-theory variance-ratio test from the supplied summaries. The R vector omits Stata confidence intervals and one-sided probability columns.
Stata reference · R reference 
signrank y = x
 
stats::wilcox.test(data$y,data$x,paired=TRUE,exact=FALSE,correct=FALSE)
stats 		Use a paired signed-rank test with ties
Differs from Stata: Uses the paired Wilcoxon asymptotic calculation without continuity correction. R omits zero differences; ties and small-sample variance conventions can affect the reference approximation.
Stata reference · R reference 
signtest y = x
 
{ d <- data$y-data$x; stats::binom.test(sum(d>0),sum(d!=0),p=0.5) }
stats 		Test the median paired difference using signs
Differs from Stata: Drops zero paired differences and applies an exact two-sided binomial test. The displayed R interval concerns the probability of a positive difference, not a confidence interval for the paired median.
Stata reference · R reference 
spearman y x
 
stats::cor.test(data$y,data$x,method="spearman",exact=FALSE)
stats 		Test Spearman rank association with ties
Differs from Stata: Uses the asymptotic Spearman test explicitly. R and Stata p-value approximations can differ, especially with ties; compare the rank correlation on the same complete pairs.
Stata reference · R reference 
swilk y
 
stats::shapiro.test(data$y)
stats 		Test univariate normality with Shapiro-Wilk
Differs from Stata: Computes the Shapiro-Wilk W test; R permits 3 to 5000 nonmissing observations. Stata also displays transformed statistics and uses its own approximation and sample-size rules.
Stata reference · R reference 
symmetry rater1 rater2
 
stats::mcnemar.test(table(data$rater1,data$rater2),correct=FALSE)
stats 		Test symmetry in a matched binary table
Differs from Stata: For a 2 by 2 table this is the uncorrected McNemar test; larger-table symmetry needs a different procedure.
Stata reference · R reference 
tabi 12 8 \ 5 15, chi2
 
stats::chisq.test(matrix(c(12,8,5,15),nrow=2,byrow=TRUE),correct=FALSE)
stats 		Run a Pearson test from immediate cell counts
Differs from Stata: Builds the stated table by rows and disables continuity correction for Pearson chi-square. R returns an htest object rather than Stata contingency-table output.
Stata reference · R reference 
tabulate g b, chi2
 
stats::chisq.test(table(data$g,data$b),correct=FALSE)
stats 		Test association in a two-way count table
Differs from Stata: table excludes missing observations and unused factor levels can affect dimensions. The R htest object provides the selected Pearson or Fisher test without Stata row and column percentage displays.
Stata reference · R reference 
tabulate g b, exact
 
stats::fisher.test(table(data$g,data$b))
stats 		Run a Fisher exact test for a two-way table
Differs from Stata: table excludes missing observations and unused factor levels can affect dimensions. The R htest object provides the selected Pearson or Fisher test without Stata row and column percentage displays.
Stata reference · R reference 
ttest y == 8
 
stats::t.test(data$y,mu=8)
stats 		Test a mean against a specified value
Differs from Stata: Selects pooled, Welch or paired inference explicitly. R returns an htest object and may reverse the reported difference when factor-level order differs; match group order and the nonmissing sample.
Stata reference · R reference 
ttest y == 8, level(90)
 
stats::t.test(data$y,mu=8,conf.level=0.90)
stats 		Compute a 90 percent interval for a mean
Differs from Stata: Selects pooled, Welch or paired inference explicitly. R returns an htest object and may reverse the reported difference when factor-level order differs; match group order and the nonmissing sample.
Stata reference · R reference 
ttest y == x
 
stats::t.test(data$y,data$x,paired=TRUE)
stats 		Test paired mean differences
Differs from Stata: Selects pooled, Welch or paired inference explicitly. R returns an htest object and may reverse the reported difference when factor-level order differs; match group order and the nonmissing sample.
Stata reference · R reference 
ttest y, by(g)
 
stats::t.test(y~g,data=data,var.equal=TRUE)
stats 		Compare two independent means with pooled variance
Differs from Stata: Selects pooled, Welch or paired inference explicitly. R returns an htest object and may reverse the reported difference when factor-level order differs; match group order and the nonmissing sample.
Stata reference · R reference 
ttest y, by(g) unequal
 
stats::t.test(y~g,data=data,var.equal=FALSE)
stats 		Compare two means using Welch unequal-variance inference
Differs from Stata: Selects pooled, Welch or paired inference explicitly. R returns an htest object and may reverse the reported difference when factor-level order differs; match group order and the nonmissing sample.
Stata reference · R reference 
ttesti 30 10 2 8
 
{ t <- (10-8)/(2/sqrt(30)); c(t=t,df=29,p.value=2*stats::pt(-abs(t),29)) }
stats 		Run a one-mean t test from summary statistics
Differs from Stata: Computes the one-sample t statistic, degrees of freedom and two-sided p-value from the supplied summaries. The R vector omits the confidence interval and one-sided probabilities.
Stata reference · R reference

Multivariate and psychometric methods

Stata R Purpose, qualifications and sources 
alpha q1 q2 q3, asis
 
psych::alpha(data[c("q1","q2","q3")],check.keys=FALSE)
psych 		Estimate raw Cronbach alpha without reversing items
Differs from Stata: asis disables Stata item-direction reversal; check.keys=FALSE makes the R direction rule explicit.
Stata reference · R reference 
alpha q1 q2 q3, generate(score) asis
 
score <- rowMeans(data[c("q1","q2","q3")])
base 		Generate a mean score from a reliability scale
Differs from Stata: asis disables Stata item-direction reversal; check.keys=FALSE makes the R direction rule explicit.
Stata reference · R reference 
alpha q1 q2 q3, item asis
 
psych::alpha(data[c("q1","q2","q3")],check.keys=FALSE)$alpha.drop
psych 		Assess reliability when each item is deleted
Differs from Stata: asis disables Stata item-direction reversal; check.keys=FALSE makes the R direction rule explicit.
Stata reference · R reference 
alpha q1 q2 q3, std asis
 
psych::alpha(data[c("q1","q2","q3")],check.keys=FALSE)$total$std.alpha
psych 		Estimate standardized Cronbach alpha
Differs from Stata: asis disables Stata item-direction reversal; check.keys=FALSE makes the R direction rule explicit.
Stata reference · R reference 
ca g h
 
MASS::corresp(unclass(table(data$g,data$h)),nf=2)
MASS 		Perform simple correspondence analysis of a contingency table
Differs from Stata: MASS returns standard coordinates; Stata ca defaults to normalize(symmetric), so plotted coordinates differ in scale. Singular values and inertias are comparable, while axes may reverse signs. The table retains category names.
Stata reference · R reference 
camat C, dimensions(2)
 
MASS::corresp(C,nf=2)
MASS 		Perform correspondence analysis of a supplied count matrix
Differs from Stata: MASS returns standard correspondence coordinates; Stata defaults to symmetric normalization. Singular values are comparable, but coordinate scaling and axis signs can differ.
Stata reference · R reference 
candisc x y z w, group(g)
 
MASS::lda(g ~ x+y+z+w,data=data,prior=rep(1/3,3))
MASS 		Estimate canonical linear discriminant directions
Differs from Stata: MASS lda supplies discriminant directions with equal class priors. This is a discriminant-analysis workflow, not the full Stata candisc canonical-test and normalization output.
Stata reference · R reference 
canon (x y) (z w)
 
stats::cancor(data[c("x","y")],data[c("z","w")])
stats 		Estimate canonical correlations between two variable sets
Differs from Stata: Returns canonical correlations and coefficient matrices for the two variable blocks. Canonical signs and coefficient normalization are arbitrary; Stata also reports significance tests.
Stata reference · R reference 
cluster averagelinkage x y z w
 
stats::hclust(stats::dist(scale(as.matrix(data[c("x","y","z","w")]))),method="average")
stats 		Fit averagelinkage hierarchical clustering on standardized variables
Differs from Stata: Standardize variables before the analogous Stata call. Ward distance and squaring conventions require care.
Stata reference · R reference 
cluster centroidlinkage x y z w
 
stats::hclust(stats::dist(scale(as.matrix(data[c("x","y","z","w")]))),method="centroid")
stats 		Fit centroidlinkage hierarchical clustering on standardized variables
Differs from Stata: Standardize variables before the analogous Stata call. Ward distance and squaring conventions require care.
Stata reference · R reference 
cluster completelinkage x y z w
 
stats::hclust(stats::dist(scale(as.matrix(data[c("x","y","z","w")]))),method="complete")
stats 		Fit completelinkage hierarchical clustering on standardized variables
Differs from Stata: Standardize variables before the analogous Stata call. Ward distance and squaring conventions require care.
Stata reference · R reference 
cluster generate group = groups(3)
 
stats::cutree(fit,k=3)
stats 		Cut a hierarchical cluster tree into three groups
Differs from Stata: Cuts the existing hclust tree at three groups. The result is an integer membership vector; cluster labels are arbitrary and Stata generates a named dataset variable.
Stata reference · R reference 
cluster kmeans x y z w, k(3)
 
stats::kmeans(scale(as.matrix(data[c("x","y","z","w")])),centers=3,nstart=10)
stats 		Partition standardized observations into three k-means clusters
Differs from Stata: Algorithms, starts and random streams differ, so cluster numbering and local optima may differ.
Stata reference · R reference 
cluster medianlinkage x y z w
 
stats::hclust(stats::dist(scale(as.matrix(data[c("x","y","z","w")]))),method="median")
stats 		Fit medianlinkage hierarchical clustering on standardized variables
Differs from Stata: Standardize variables before the analogous Stata call. Ward distance and squaring conventions require care.
Stata reference · R reference 
cluster singlelinkage x y z w
 
stats::hclust(stats::dist(scale(as.matrix(data[c("x","y","z","w")]))),method="single")
stats 		Fit singlelinkage hierarchical clustering on standardized variables
Differs from Stata: Standardize variables before the analogous Stata call. Ward distance and squaring conventions require care.
Stata reference · R reference 
cluster wardslinkage x y z w
 
stats::hclust(stats::dist(scale(as.matrix(data[c("x","y","z","w")]))),method="ward.D2")
stats 		Fit wardslinkage hierarchical clustering on standardized variables
Differs from Stata: Standardize variables before the analogous Stata call. Ward distance and squaring conventions require care.
Stata reference · R reference 
cluster waveragelinkage x y z w
 
stats::hclust(stats::dist(scale(as.matrix(data[c("x","y","z","w")]))),method="mcquitty")
stats 		Fit waveragelinkage hierarchical clustering on standardized variables
Differs from Stata: Standardize variables before the analogous Stata call. Ward distance and squaring conventions require care.
Stata reference · R reference 
clustermat completelinkage D
 
stats::hclust(stats::as.dist(D),method="complete")
stats 		Cluster a precomputed dissimilarity matrix
Differs from Stata: Uses complete linkage on the supplied dissimilarity matrix. Dissimilarity definitions and tie breaking affect the tree; R returns an hclust object rather than a Stata named cluster solution.
Stata reference · R reference 
discrim knn x y z w, group(g) k(5)
 
class::knn(train=as.matrix(data[c("x","y","z","w")]),test=as.matrix(data[c("x","y","z","w")]),cl=data$g,k=5)
class 		Classify observations using five nearest neighbors
Differs from Stata: Selects predictors by name and classifies the training observations. class::knn uses unweighted majority voting and random tie breaking; Stata defaults to equal class priors, which can change decisions for unbalanced classes. Use held-out data for predictive accuracy.
Stata reference · R reference 
discrim lda x y z w, group(g)
 
MASS::lda(g ~ x+y+z+w,data=data,prior=rep(1/3,3))
MASS 		Fit LDA discriminant analysis
Differs from Stata: Prior class probabilities and covariance estimation conventions should be specified before comparing results.
Stata reference · R reference 
discrim lda x y z w, group(g) lootable
 
MASS::lda(g ~ x+y+z+w,data=data,CV=TRUE,prior=rep(1/3,3))
MASS 		Classify using leave-one-out linear discriminant analysis
Stata lootable requests leave-one-out classification; MASS CV=TRUE performs the corresponding prediction workflow.
Stata reference · R reference 
discrim lda x y z w, group(g) priors(equal)
 
MASS::lda(g ~ x+y+z+w,data=data,prior=rep(1/3,3))
MASS 		Use equal prior class probabilities in LDA
Differs from Stata: Equal class priors are explicit where shown. MASS coefficient normalization and class order can differ from Stata; compare posterior classifications after aligning priors and covariance estimates.
Stata reference · R reference 
discrim qda x y z w, group(g)
 
{ analysis <- droplevels(stats::model.frame(g ~ x + y + z + w, data = data, na.action = stats::na.omit)); groups <- length(unique(analysis$g)); MASS::qda(g ~ x + y + z + w, data = analysis, prior = rep(1 / groups, groups)) }
MASS 		Fit QDA discriminant analysis
Differs from Stata: Explicit equal priors are computed from classes remaining after complete-case filtering of the response and predictors, matching Stata discrim qda default priors(equal). MASS otherwise uses sample class proportions. Align covariance normalization and class labels before comparing posterior probabilities.
Stata reference · R reference 
estat gof, stats(all)
 
lavaan::fitMeasures(fit,c("chisq","df","cfi","tli","rmsea","srmr"))
lavaan 		Report confirmatory-factor model fit indices
Differs from Stata: Selects six lavaan SEM fit indices. Model identification, likelihood scaling, missing-data estimation and baseline-model definitions must match; this is not every statistic in Stata stats(all).
Stata reference · R reference 
estat kmo
 
psych::KMO(stats::cor(X))
psych 		Assess factorability using Kaiser-Meyer-Olkin measures
Differs from Stata: Computes KMO measures from the supplied correlation matrix. R returns overall and per-variable measures in a list; match the same correlation and missing-data sample used for Stata factor.
Stata reference · R reference 
estat mindices
 
lavaan::modindices(fit)
lavaan 		Inspect SEM modification indices
Differs from Stata: Returns lavaan modification indices for admissible freed parameters. Align model constraints, information matrix and missing-data likelihood; candidate parameter sets can differ from Stata.
Stata reference · R reference 
estat residuals
 
lavaan::lavResiduals(fit,type="raw")
lavaan 		Inspect residual correlations after SEM
Differs from Stata: Uses raw observed-minus-implied covariance residuals, matching Stata default raw residuals. lavaan also returns standardized summaries with its own scaling, which are not the same Stata output columns.
Stata reference · R reference 
estat smc
 
1-1/diag(solve(stats::cor(X)))
stats 		Compute squared multiple correlations for scale items
Differs from Stata: Computes squared multiple correlations as one minus the reciprocal diagonal of the inverse correlation matrix. Requires a nonsingular correlation matrix and returns a vector rather than Stata display tables.
Stata reference · R reference 
factor q1 q2 q3 q4 q5 q6, factors(1) ml
 
stats::factanal(as.matrix(data[c("q1","q2","q3","q4","q5","q6")]),factors=1,rotation="none")
stats 		Fit a one-factor maximum-likelihood factor model
Differs from Stata: Matches the stated ML or principal-component extraction and factor count. Loading signs and rotations are indeterminate; psych principal-component scores and scaling need not reproduce Stata factor postestimation.
Stata reference · R reference 
factor q1 q2 q3 q4 q5 q6, factors(2) ipf
 
psych::fa(as.matrix(data[c("q1","q2","q3","q4","q5","q6")]),nfactors=2,fm="pa",rotate="none")
psych 		Fit iterated principal-axis factors
Differs from Stata: Principal-axis algorithms and convergence tolerances vary; do not assume identical loadings.
Stata reference · R reference 
factor q1 q2 q3 q4 q5 q6, factors(2) ml
 
stats::factanal(as.matrix(data[c("q1","q2","q3","q4","q5","q6")]),factors=2,rotation="none")
stats 		Fit a two-factor maximum-likelihood factor model
Differs from Stata: Matches the stated ML or principal-component extraction and factor count. Loading signs and rotations are indeterminate; psych principal-component scores and scaling need not reproduce Stata factor postestimation.
Stata reference · R reference 
factor q1 q2 q3 q4 q5 q6, factors(2) pcf
 
psych::principal(as.matrix(data[c("q1","q2","q3","q4","q5","q6")]),nfactors=2,rotate="none")
psych 		Fit a principal-component factor solution
Differs from Stata: Matches the stated ML or principal-component extraction and factor count. Loading signs and rotations are indeterminate; psych principal-component scores and scaling need not reproduce Stata factor postestimation.
Stata reference · R reference 
correlate q1 q2 q3 q4 q5 q6
matrix R = r(C)
factormat R, n(300) factors(2) ml
 
stats::factanal(covmat=R,n.obs=300,factors=2,rotation="none")
stats 		Fit maximum-likelihood factors from a correlation matrix
Differs from Stata: Fixed workflow. R is a fixed correlation-matrix object; sample size supplies likelihood-based fit statistics.
Stata reference · R reference 
hotelling x y
 
{ n <- nrow(data); k <- 2; m <- colMeans(data[c("x","y")]); T2 <- n*drop(t(m)%*%solve(stats::cov(data[c("x","y")]),m)); F <- (n-k)/(k*(n-1))*T2; c(T2=T2,F=F,p.value=stats::pf(F,k,n-k,lower.tail=FALSE)) }
stats 		Test a vector of means against zero with Hotelling T squared
Differs from Stata: Computes the one-sample Hotelling T-squared and its F transformation for these two variables. Assumes a complete sample and invertible covariance; Stata additionally displays mean summaries.
Stata reference · R reference 
kap rater1 rater2
 
psych::cohen.kappa(table(data$rater1,data$rater2))$kappa
psych 		Estimate unweighted Cohen kappa for two raters
Differs from Stata: Returns unweighted Cohen kappa from the cross-tabulation. It omits Stata agreement summaries, standard errors and confidence intervals; align missing-value and category handling.
Stata reference · R reference 
manova y z = g
 
summary(stats::manova(cbind(y,z)~g,data=data),test="Pillai")
stats 		Test multivariate group differences using Pillai
Differs from Stata: Stata displays all four MANOVA statistics; this R example selects the named statistic.
Stata reference · R reference 
manova y z = g c.x
 
stats::manova(cbind(y,z)~g+x,data=data)
stats 		Fit multivariate analysis of covariance
Differs from Stata: Fits the stated multivariate linear model. R summary defaults to Pillai while Stata displays several multivariate criteria; choose the same criterion, hypothesis and factor contrasts.
Stata reference · R reference 
manova y z = g
 
summary(stats::manova(cbind(y,z)~g,data=data),test="Wilks")
stats 		Test multivariate group differences using Wilks
Differs from Stata: Stata displays all four MANOVA statistics; this R example selects the named statistic.
Stata reference · R reference 
manova y z = g
 
summary(stats::manova(cbind(y,z)~g,data=data),test="Hotelling-Lawley")
stats 		Test multivariate group differences using Hotelling-Lawley
Differs from Stata: Stata displays all four MANOVA statistics; this R example selects the named statistic.
Stata reference · R reference 
manova y z = g
 
summary(stats::manova(cbind(y,z)~g,data=data),test="Roy")
stats 		Test multivariate group differences using Roy
Differs from Stata: Stata displays all four MANOVA statistics; this R example selects the named statistic.
Stata reference · R reference 
mca g h j, method(indicator)
 
MASS::mca(data, nf=2)
MASS 		Analyze multiple categorical variables through indicator MCA
Differs from Stata: MASS uses an indicator-matrix MCA; inertia normalization differs from adjusted and joint MCA.
Stata reference · R reference 
mds x y z w, id(id) dimension(2) noplot
 
stats::cmdscale(stats::dist(as.matrix(data[c("x","y","z","w")])),k=2,eig=TRUE)
stats 		Fit classical metric multidimensional scaling
Differs from Stata: Uses classical metric scaling of Euclidean distances. Stata mds defaults and stress criteria can differ; coordinates are comparable only after aligning distances, method, rotation and reflection.
Stata reference · R reference 
mds x y z w, id(id) method(nonmetric) std noplot
 
MASS::isoMDS(stats::dist(scale(as.matrix(data[c("x","y","z","w")]))),k=2,trace=FALSE)
MASS 		Fit nonmetric multidimensional scaling
Differs from Stata: Nonmetric optimization, stress definitions and starting configurations can differ.
Stata reference · R reference 
mdsmat D, dimension(2) noplot
 
stats::cmdscale(stats::as.dist(D),k=2)
stats 		Scale an existing dissimilarity matrix
Differs from Stata: Applies classical metric scaling to the supplied dissimilarity matrix. Match Stata metric method and eigenvalue treatment; coordinate rotations and reflections are arbitrary.
Stata reference · R reference 
mvreg y z = x w
 
stats::lm(cbind(y,z) ~ x+w,data=data)
stats 		Fit multivariate multiple regression
Differs from Stata: Returns a multivariate lm fit on the common complete-case sample. Stata mvreg displays separate equations and cross-equation summaries; request the corresponding R summaries explicitly.
Stata reference · R reference 
mvtest normality x y z w, stats(skewness kurtosis)
 
psych::mardia(X,plot=FALSE)
psych 		Compute Mardia multivariate skewness and kurtosis tests
Differs from Stata: Finite-sample corrections may differ; this selects Mardia components rather than Stata default Doornik-Hansen.
Stata reference · R reference 
pca x y z w
 
stats::prcomp(as.matrix(data[c("x","y","z","w")]),center=TRUE,scale.=TRUE)
stats 		Fit correlation-matrix PCA
Differs from Stata: Selects correlation or covariance PCA explicitly. Component signs are arbitrary; prcomp returns singular-value-based objects and scores whose scaling differs from Stata displays.
Stata reference · R reference 
pca x y z w, components(2)
 
stats::prcomp(as.matrix(data[c("x","y","z","w")]),scale.=TRUE,rank.=2)
stats 		Retain two principal components
Differs from Stata: Selects correlation or covariance PCA explicitly. Component signs are arbitrary; prcomp returns singular-value-based objects and scores whose scaling differs from Stata displays.
Stata reference · R reference 
pca x y z w, covariance
 
stats::prcomp(as.matrix(data[c("x","y","z","w")]),center=TRUE,scale.=FALSE)
stats 		Fit covariance-matrix PCA
Differs from Stata: Selects correlation or covariance PCA explicitly. Component signs are arbitrary; prcomp returns singular-value-based objects and scores whose scaling differs from Stata displays.
Stata reference · R reference 
correlate x y z w
matrix R = r(C)
pcamat R, n(50)
 
stats::princomp(covmat=R)
stats 		Perform PCA directly on a correlation matrix
Differs from Stata: Fixed workflow.
Stata reference · R reference 
predict p1 p2 p3, pr
 
data[c("p1","p2","p3")] <- stats::predict(fit)$posterior
MASS 		Predict posterior probabilities for all discriminant classes
Differs from Stata: Uses the existing discriminant fit and its class order. Returned labels or posterior-probability columns must match the fitted classes; R factor order need not match Stata category codes.
Stata reference · R reference 
predict pc1 pc2, score
 
data[c("pc1","pc2")] <- stats::predict(fit)[,1:2,drop=FALSE]
stats 		Compute the first two principal-component scores
Differs from Stata: Returns scores from the existing PCA fit. R score scaling and arbitrary component signs can differ from Stata predict, score; align normalization before comparing values.
Stata reference · R reference 
predict predicted, classification
 
data$predicted <- stats::predict(fit)$class
MASS 		Predict discriminant-analysis class membership
Differs from Stata: Uses the existing discriminant fit and its class order. Returned labels or posterior-probability columns must match the fitted classes; R factor order need not match Stata category codes.
Stata reference · R reference 
predict score1 score2, bartlett
 
data[c("score1","score2")] <- stats::factanal(as.matrix(data[c("q1","q2","q3","q4","q5","q6")]),factors=2,rotation="none",scores="Bartlett")$scores
stats 		Compute Bartlett-method factor scores
Differs from Stata: Fixed measurement model with indicators q1-q6; output score names can be changed.
Stata reference · R reference 
predict score1 score2, regression
 
data[c("score1","score2")] <- stats::factanal(as.matrix(data[c("q1","q2","q3","q4","q5","q6")]),factors=2,rotation="none",scores="regression")$scores
stats 		Compute regression-method factor scores
Differs from Stata: Fixed measurement model with indicators q1-q6; output score names can be changed.
Stata reference · R reference 
rotate, oblique oblimin
 
GPArotation::oblimin(unclass(fit$loadings))
GPArotation 		Use the oblique oblimin loading-rotation criterion
Differs from Stata: Rotates the supplied loading matrix only. Loading signs, axis order and normalization can differ from Stata; correlated-factor solutions also require comparing the factor-correlation matrix.
Stata reference · R reference 
rotate, orthogonal quartimax
 
GPArotation::quartimax(unclass(fit$loadings))
GPArotation 		Use the orthogonal quartimax loading-rotation criterion
Differs from Stata: Rotates the supplied loading matrix only. Loading signs, axis order and normalization can differ from Stata; correlated-factor solutions also require comparing the factor-correlation matrix.
Stata reference · R reference 
rotate, promax
 
stats::promax(fit$loadings,m=3)
stats 		Rotate factor loadings using promax
Differs from Stata: Loadings are identified only up to sign and ordering; rotation normalization conventions may differ.
Stata reference · R reference 
rotate, varimax
 
stats::varimax(fit$loadings,normalize=FALSE)
stats 		Rotate factor loadings using varimax
Differs from Stata: Loadings are identified only up to sign and ordering; rotation normalization conventions may differ.
Stata reference · R reference 
rotatemat L, varimax
 
stats::varimax(L)
stats 		Rotate an externally supplied loading matrix
Differs from Stata: Applies varimax to the supplied matrix. Match normalization and column ordering; equivalent rotated solutions may reverse column signs.
Stata reference · R reference 
sem (visual -> x1 x2 x3) (textual -> x4 x5 x6) (speed -> x7 x8 x9), latent(visual textual speed)
 
lavaan::cfa("visual =~ x1+x2+x3; textual =~ x4+x5+x6; speed =~ x7+x8+x9",data=data)
lavaan 		Fit a three-factor confirmatory measurement model
Differs from Stata: Latent-variable scale constraints and mean structures must be aligned when comparing estimates.
Stata reference · R reference 
sem, standardized
 
lavaan::standardizedSolution(fit)
lavaan 		Report standardized SEM parameters
Differs from Stata: Fixed workflow.
Stata reference · R reference 
tetrachoric a b
 
polycor::polychor(data$a,data$b,ML=TRUE,std.err=TRUE)
polycor 		Estimate a tetrachoric correlation by maximum likelihood
Differs from Stata: Each binary variable must have two observed levels; inference assumes latent bivariate normality.
Stata reference · R reference

Power and precision planning

Stata R Purpose, qualifications and sources 
ciwidth onemean, sd(2) n(50)
 
2*stats::qt(0.975,df=49)*2/sqrt(50)
stats 		Calculate the anticipated full width of a mean interval
Anticipated-width workflow substituting the design SD for the future sample SD; interval width is upper minus lower. Stata also supports probability-of-width planning, which is not implemented here.
Stata reference · R reference 
ciwidth onemean, sd(2) width(1)
 
ceiling(stats::uniroot(function(n) 2*stats::qt(0.975,n-1)*2/sqrt(n)-1,c(3,1000))$root)
stats 		Plan n for a target mean-interval width
Anticipated-width workflow substituting the design SD for the future sample SD; interval width is upper minus lower. Stata also supports probability-of-width planning, which is not implemented here.
Stata reference · R reference 
ciwidth pairedmeans, sddiff(2) n(50)
 
2*stats::qt(0.975,49)*2/sqrt(50)
stats 		Calculate a paired mean-difference interval width
Anticipated-width workflow substituting the design SD for the future sample SD; interval width is upper minus lower. Stata also supports probability-of-width planning, which is not implemented here.
Stata reference · R reference 
ciwidth twomeans, sd(2) n1(40) n2(60)
 
2*stats::qt(0.975,98)*2*sqrt(1/40+1/60)
stats 		Calculate an equal-variance difference-of-means interval width
Anticipated-width workflow substituting the design SD for the future sample SD; interval width is upper minus lower. Stata also supports probability-of-width planning, which is not implemented here.
Stata reference · R reference 
power onecorrelation 0 0.3, n(100)
 
pwr::pwr.r.test(r=0.3,n=100,sig.level=0.05)
pwr 		Calculate correlation-test power at fixed n
Differs from Stata: pwr uses its documented approximation and effect-size parameterization; numerical results need not equal Stata.
Stata reference · R reference 
power onecorrelation 0 0.3, power(0.8)
 
pwr::pwr.r.test(r=0.3,power=0.8,sig.level=0.05)
pwr 		Plan sample size for detecting a correlation
Differs from Stata: pwr uses its documented approximation and effect-size parameterization; numerical results need not equal Stata.
Stata reference · R reference 
power onecorrelation 0, n(100) power(0.8)
 
pwr::pwr.r.test(r=NULL,n=100,power=0.8,sig.level=0.05)
pwr 		Find a detectable correlation at fixed n
Differs from Stata: pwr uses its documented approximation and effect-size parameterization; numerical results need not equal Stata.
Stata reference · R reference 
power onemean 0 0.5, sd(1) knownsd n(40)
 
pwr::pwr.norm.test(d=0.5,n=40,sig.level=0.05)
pwr 		Calculate one-mean z-test power with known SD
Differs from Stata: pwr reports the mean difference in SD units; knownsd distinguishes this from a t test.
Stata reference · R reference 
power onemean 0 0.5, sd(1) knownsd power(0.8)
 
pwr::pwr.norm.test(d=0.5,power=0.8,sig.level=0.05)
pwr 		Plan a one-mean z-test sample size with known SD
Differs from Stata: pwr reports the mean difference in SD units; knownsd distinguishes this from a t test.
Stata reference · R reference 
power onemean 0 0.5, sd(1) n(40)
 
stats::power.t.test(n=40,delta=0.5,sd=1,type="one.sample",sig.level=0.05)
stats 		Calculate power for a fixed-size onemean t test
Differs from Stata: Computes the displayed normal-theory t-test or two-proportion power design. R two-sample n is per group, whereas Stata n() is total; alternative, allocation and strict two-sided rejection conventions must match.
Stata reference · R reference 
power onemean 0 0.5, sd(1) n(40) onesided
 
stats::power.t.test(n=40,delta=0.5,sd=1,type="one.sample",alternative="one.sided")
stats 		Calculate one-sided one-mean t-test power
Differs from Stata: Computes the displayed normal-theory t-test or two-proportion power design. R two-sample n is per group, whereas Stata n() is total; alternative, allocation and strict two-sided rejection conventions must match.
Stata reference · R reference 
power onemean 0 0.5, sd(1) power(0.8)
 
stats::power.t.test(delta=0.5,sd=1,type="one.sample",sig.level=0.05,power=0.8)
stats 		Plan the sample size for a onemean t test
Differs from Stata: R reports n per group for an independent two-sample test; round sample sizes upward.
Stata reference · R reference 
power onemean 0, sd(1) knownsd n(40) power(0.8)
 
pwr::pwr.norm.test(d=NULL,n=40,power=0.8,sig.level=0.05)
pwr 		Find a detectable one-mean z-test effect
Differs from Stata: pwr reports the mean difference in SD units; knownsd distinguishes this from a t test.
Stata reference · R reference 
power onemean 0, sd(1) n(40) power(0.8)
 
stats::power.t.test(n=40,delta=NULL,sd=1,power=0.8,type="one.sample")
stats 		Find the detectable difference for a onemean t test
Differs from Stata: R returns a positive detectable difference; add it to the null mean to recover the alternative mean.
Stata reference · R reference 
power oneproportion 0.3 0.5, n(100)
 
pwr::pwr.p.test(h=pwr::ES.h(0.5,0.3),n=100,sig.level=0.05)
pwr 		Approximate one-proportion power for a fixed sample
Uses the arcsine effect h, an alternative to Stata default score-test calculations.
Stata reference · R reference 
power oneproportion 0.3 0.5, power(0.8)
 
pwr::pwr.p.test(h=pwr::ES.h(0.5,0.3),power=0.8,sig.level=0.05)
pwr 		Approximate a one-proportion design using Cohen h
Uses the arcsine effect h, an alternative to Stata default score-test calculations.
Stata reference · R reference 
power oneway 0 0.5 1, varerror(1) n(90)
 
pwr::pwr.anova.test(k=3,n=30,f=sqrt(mean((c(0,0.5,1)-0.5)^2)))
pwr 		Calculate balanced one-way ANOVA power
Differs from Stata: pwr uses its documented approximation and effect-size parameterization; numerical results need not equal Stata.
Stata reference · R reference 
power oneway 0 0.5 1, varerror(1) power(0.8)
 
pwr::pwr.anova.test(k=3,f=sqrt(mean((c(0,0.5,1)-0.5)^2)),power=0.8)
pwr 		Plan balanced one-way ANOVA group size
Differs from Stata: pwr uses its documented approximation and effect-size parameterization; numerical results need not equal Stata.
Stata reference · R reference 
power pairedmeans 0 0.5, sddiff(1) n(40)
 
stats::power.t.test(n=40,delta=0.5,sd=1,type="paired",sig.level=0.05)
stats 		Calculate power for a fixed-size pairedmeans t test
Differs from Stata: Computes the displayed normal-theory t-test or two-proportion power design. R two-sample n is per group, whereas Stata n() is total; alternative, allocation and strict two-sided rejection conventions must match.
Stata reference · R reference 
power pairedmeans 0 0.5, sddiff(1) power(0.8)
 
stats::power.t.test(delta=0.5,sd=1,type="paired",sig.level=0.05,power=0.8)
stats 		Plan the sample size for a pairedmeans t test
Differs from Stata: R reports n per group for an independent two-sample test; round sample sizes upward.
Stata reference · R reference 
power pairedmeans 0, sddiff(1) n(40) power(0.8)
 
stats::power.t.test(n=40,delta=NULL,sd=1,power=0.8,type="paired")
stats 		Find the detectable difference for a pairedmeans t test
Differs from Stata: R returns a positive detectable difference; add it to the null mean to recover the alternative mean.
Stata reference · R reference 
power rsquared 0.1, n(100) ntested(3)
 
pwr::pwr.f2.test(u=3,v=96,f2=0.1/(1-0.1))
pwr 		Calculate power for an overall three-predictor R squared test
Differs from Stata: f2=(R2_full-R2_reduced)/(1-R2_full); residual df is N minus all predictors minus one.
Stata reference · R reference 
power rsquared 0.1, power(0.8) ntested(3)
 
{ design <- pwr::pwr.f2.test(u=3,f2=0.1/(1-0.1),power=0.8); ceiling(design$v+3+1) }
pwr 		Plan total n for an overall R squared test
Differs from Stata: f2=(R2_full-R2_reduced)/(1-R2_full); residual df is N minus all predictors minus one.
Stata reference · R reference 
power rsquared 0.2 0.3, n(120) ntested(2) ncontrol(3)
 
pwr::pwr.f2.test(u=2,v=114,f2=(0.3-0.2)/(1-0.3))
pwr 		Calculate power for incremental explained variance
Differs from Stata: f2=(R2_full-R2_reduced)/(1-R2_full); residual df is N minus all predictors minus one.
Stata reference · R reference 
power twomeans 0 0.5, sd(1) n(80)
 
stats::power.t.test(n=40.0,delta=0.5,sd=1,type="two.sample",sig.level=0.05)
stats 		Calculate power for a fixed-size twomeans t test
Differs from Stata: Computes the displayed normal-theory t-test or two-proportion power design. R two-sample n is per group, whereas Stata n() is total; alternative, allocation and strict two-sided rejection conventions must match.
Stata reference · R reference 
power twomeans 0 0.5, sd(1) n1(30) n2(50)
 
pwr::pwr.t2n.test(n1=30,n2=50,d=0.5,sig.level=0.05)
pwr 		Calculate power with unequal independent group sizes
Differs from Stata: Uses the pooled-variance standardized mean-difference model; this is not a Welch heteroskedastic design.
Stata reference · R reference 
power twomeans 0 0.5, sd(1) power(0.8)
 
stats::power.t.test(delta=0.5,sd=1,type="two.sample",sig.level=0.05,power=0.8)
stats 		Plan the sample size for a twomeans t test
Differs from Stata: R reports n per group for an independent two-sample test; round sample sizes upward.
Stata reference · R reference 
power twomeans 0, sd(1) n(80) power(0.8)
 
stats::power.t.test(n=40.0,delta=NULL,sd=1,power=0.8,type="two.sample")
stats 		Find the detectable difference for a twomeans t test
Differs from Stata: R returns a positive detectable difference; add it to the null mean to recover the alternative mean.
Stata reference · R reference 
power twoproportions 0.3 0.5, n(200)
 
stats::power.prop.test(n=100,p1=0.3,p2=0.5,sig.level=0.05)
stats 		Calculate power for two fixed-size proportion samples
Differs from Stata: Computes the displayed normal-theory t-test or two-proportion power design. R two-sample n is per group, whereas Stata n() is total; alternative, allocation and strict two-sided rejection conventions must match.
Stata reference · R reference 
power twoproportions 0.3 0.5, n(200) onesided
 
stats::power.prop.test(n=100,p1=0.3,p2=0.5,alternative="one.sided")
stats 		Calculate one-sided two-proportion power
Differs from Stata: Computes the displayed normal-theory t-test or two-proportion power design. R two-sample n is per group, whereas Stata n() is total; alternative, allocation and strict two-sided rejection conventions must match.
Stata reference · R reference 
power twoproportions 0.3 0.5, power(0.8)
 
stats::power.prop.test(p1=0.3,p2=0.5,power=0.8,sig.level=0.05)
stats 		Plan sample size for two independent proportions
Differs from Stata: R reports sample size per group. Normal-approximation conventions and continuity corrections may differ.
Stata reference · R reference 
power twoproportions 0.3, n(200) power(0.8)
 
stats::power.prop.test(n=100,p1=0.3,p2=NULL,power=0.8,sig.level=0.05)
stats 		Find a detectable alternative proportion
Differs from Stata: Computes the displayed normal-theory t-test or two-proportion power design. R two-sample n is per group, whereas Stata n() is total; alternative, allocation and strict two-sided rejection conventions must match.
Stata reference · R reference

Resampling and simulation

Stata R Purpose, qualifications and sources 
bayesboot _b[x], reps(199): regress y x
 
replicate(199,{ weights <- stats::rexp(nrow(data)); unname(stats::coef(stats::lm(y~x,data=data,weights=weights))["x"]) })
stats 		Use Bayesian-bootstrap Dirichlet weights for a slope
Fixed workflow. Independent exponential weights normalized to sum one are Dirichlet(1,...,1); common scaling does not change weighted least-squares coefficients.
Stata reference · R reference 
bootstrap _b, reps(199) strata(g): regress y x
 
boot::boot(data,statistic=function(d,i) stats::coef(stats::lm(y~x,data=d[i,])),R=199,strata=data$g)
boot 		Bootstrap regression coefficients within fixed strata
Fixed workflow. Random streams and resampling defaults differ across software; a fixed R seed makes this workflow reproducible.
Stata reference · R reference 
bootstrap difference=(_b[x]-_b[z]), reps(199): regress y x z
 
boot::boot(data,statistic=function(d,i) unname(diff(stats::coef(stats::lm(y~x+z,data=d[i,]))[c("z","x")])),R=199)
boot 		Bootstrap a difference between regression slopes
Fixed workflow. Random streams and resampling defaults differ across software; a fixed R seed makes this workflow reproducible.
Stata reference · R reference 
bootstrap mean=r(mean), reps(199) cluster(g): summarize y
 
{ ids <- unique(data$g); replicate(199,{ picked <- sample(ids,length(ids),replace=TRUE); sample_data <- do.call(rbind,lapply(picked,function(id) data[data$g==id,,drop=FALSE])); mean(sample_data$y) }) }
base 		Bootstrap a mean by resampling complete clusters
Fixed workflow. Resamples cluster identifiers, retaining every observation in each selected cluster. Three clusters are only a demonstration.
Stata reference · R reference 
bootstrap mean=r(mean), reps(199): summarize y
 
boot::boot(data,statistic=function(d,i) mean(d$y[i]),R=199)
boot 		Bootstrap the sample mean
Fixed workflow. Random streams and resampling defaults differ across software; a fixed R seed makes this workflow reproducible.
Stata reference · R reference 
bootstrap mean=r(mean), reps(499) bca: summarize y
 
{ boot_fit <- boot::boot(data,function(d,i) mean(d$y[i]),R=499); boot::boot.ci(boot_fit,type="bca") }
boot 		Compute a bias-corrected accelerated bootstrap interval
Fixed workflow. Random streams and resampling defaults differ across software; a fixed R seed makes this workflow reproducible.
Stata reference · R reference 
bootstrap median=r(p50), reps(199): summarize y, detail
 
boot::boot(data,statistic=function(d,i) stats::median(d$y[i]),R=199)
boot 		Bootstrap the sample median
Fixed workflow. Random streams and resampling defaults differ across software; a fixed R seed makes this workflow reproducible.
Stata reference · R reference 
bootstrap q1=r(p25), reps(199): summarize y, detail
 
boot::boot(data,statistic=function(d,i) unname(stats::quantile(d$y[i],0.25,type=2)),R=199)
boot 		Bootstrap the lower quartile
Fixed workflow. Random streams and resampling defaults differ across software; a fixed R seed makes this workflow reproducible.
Stata reference · R reference 
bootstrap q3=r(p75), reps(199): summarize y, detail
 
boot::boot(data,statistic=function(d,i) unname(stats::quantile(d$y[i],0.75,type=2)),R=199)
boot 		Bootstrap the upper quartile
Fixed workflow. Random streams and resampling defaults differ across software; a fixed R seed makes this workflow reproducible.
Stata reference · R reference 
bootstrap sd=r(sd), reps(199): summarize y
 
boot::boot(data,statistic=function(d,i) stats::sd(d$y[i]),R=199)
boot 		Bootstrap the sample standard deviation
Fixed workflow. Random streams and resampling defaults differ across software; a fixed R seed makes this workflow reproducible.
Stata reference · R reference 
bootstrap slope=_b[x], reps(199): regress y x
 
boot::boot(data,statistic=function(d,i) unname(stats::coef(stats::lm(y~x,data=d[i,]))["x"]),R=199)
boot 		Bootstrap a regression slope by resampling rows
Fixed workflow. Random streams and resampling defaults differ across software; a fixed R seed makes this workflow reproducible.
Stata reference · R reference 
bootstrap variance=r(Var), reps(199): summarize y
 
boot::boot(data,statistic=function(d,i) stats::var(d$y[i]),R=199)
boot 		Bootstrap the sample variance
Fixed workflow. Random streams and resampling defaults differ across software; a fixed R seed makes this workflow reproducible.
Stata reference · R reference 
bsample
 
data[sample.int(nrow(data),nrow(data),replace=TRUE),,drop=FALSE]
base 		Draw a nonparametric row-bootstrap sample
Fixed workflow. Random streams and resampling defaults differ across software; a fixed R seed makes this workflow reproducible.
Stata reference · R reference 
bsample, strata(g)
 
do.call(rbind,lapply(split(data,data$g),function(d) d[sample.int(nrow(d),nrow(d),replace=TRUE),,drop=FALSE]))
base 		Draw a stratified row-bootstrap sample
Fixed workflow. Random streams and resampling defaults differ across software; a fixed R seed makes this workflow reproducible.
Stata reference · R reference 
bstat
 
c(estimate=boot_fit$t0,bias=mean(boot_fit$t)-boot_fit$t0,se=stats::sd(boot_fit$t))
stats 		Summarize stored bootstrap estimates, bias and standard error
Random streams and resampling defaults differ across software; a fixed R seed makes this workflow reproducible.
Stata reference · R reference 
estat bootstrap, normal
 
boot::boot.ci(boot_fit,type="norm")
boot 		Compute a normal bootstrap interval
Random streams and resampling defaults differ across software; a fixed R seed makes this workflow reproducible.
Stata reference · R reference 
estat bootstrap, percentile
 
boot::boot.ci(boot_fit,type="perc")
boot 		Compute a percentile bootstrap interval
Random streams and resampling defaults differ across software; a fixed R seed makes this workflow reproducible.
Stata reference · R reference 
jackknife _b[x]: regress y x
 
{ n <- nrow(data); deleted <- vapply(seq_len(n),function(i) unname(stats::coef(stats::lm(y~x,data=data[-i,]))["x"]),numeric(1)); c(estimate=unname(stats::coef(stats::lm(y~x,data=data))["x"]),se=sqrt((n-1)/n*sum((deleted-mean(deleted))^2))) }
stats 		Jackknife a regression slope
Fixed workflow. The delete-one jackknife is deterministic; its variance uses (n-1)/n times the squared deviations of the replicate estimates.
Stata reference · R reference 
jackknife mean=r(mean), rclass cluster(g): summarize y
 
{ groups <- unique(data$g); deleted <- vapply(groups,function(id) mean(data$y[data$g!=id]),numeric(1)); length(groups) }
base 		Enumerate delete-one-cluster jackknife replicates
Fixed workflow. Returns replicate count and retains deleted estimates. Unequal cluster-size variance corrections need an explicitly chosen jackknife design.
Stata reference · R reference 
jackknife mean=r(mean), rclass keep: summarize y
 
{ n <- nrow(data); deleted <- vapply(seq_len(n),function(i) mean(data$y[-i]),numeric(1)); n*mean(data$y)-(n-1)*deleted }
base 		Generate jackknife pseudovalues for the sample mean
Fixed workflow. The delete-one jackknife is deterministic; its variance uses (n-1)/n times the squared deviations of the replicate estimates.
Stata reference · R reference 
jackknife mean=r(mean), rclass: summarize y
 
{ n <- nrow(data); deleted <- vapply(seq_len(n),function(i) mean(data$y[-i]),numeric(1)); c(estimate=mean(data$y),se=sqrt((n-1)/n*sum((deleted-mean(deleted))^2))) }
stats 		Jackknife the sample mean
Fixed workflow. The delete-one jackknife is deterministic; its variance uses (n-1)/n times the squared deviations of the replicate estimates.
Stata reference · R reference 
jackknife sd=r(sd), rclass: summarize y
 
{ n <- nrow(data); deleted <- vapply(seq_len(n),function(i) stats::sd(data$y[-i]),numeric(1)); c(estimate=stats::sd(data$y),se=sqrt((n-1)/n*sum((deleted-mean(deleted))^2))) }
stats 		Jackknife the sample standard deviation
Fixed workflow. The delete-one jackknife is deterministic; its variance uses (n-1)/n times the squared deviations of the replicate estimates.
Stata reference · R reference 
permute y _b[x], reps(199) strata(g): regress y x
 
replicate(199,{ d <- data; d$y <- unsplit(lapply(split(data$y,data$g),sample),data$g); unname(stats::coef(stats::lm(y~x,data=d))["x"]) })
stats 		Permute outcomes within fixed strata
Fixed workflow. Random streams and resampling defaults differ across software; a fixed R seed makes this workflow reproducible.
Stata reference · R reference 
permute y _b[x], reps(199): regress y x
 
replicate(199,{ d <- data; d$y <- sample(d$y); unname(stats::coef(stats::lm(y~x,data=d))["x"]) })
stats 		Generate a permutation distribution for a regression slope
Fixed workflow. Random streams and resampling defaults differ across software; a fixed R seed makes this workflow reproducible.
Stata reference · R reference 
permute y r(rho), reps(199): correlate y x
 
replicate(199,stats::cor(sample(data$y),data$x))
stats 		Generate a permutation distribution for a correlation
Fixed workflow. Random streams and resampling defaults differ across software; a fixed R seed makes this workflow reproducible.
Stata reference · R reference 
program define mymean, rclass
    preserve
    clear
    set obs 40
    generate y = rnormal(2,1)
    summarize y
    local estimate = r(mean)
    restore
    return scalar mean = `estimate'
end
simulate mean=r(mean), reps(199): mymean
 
replicate(199,mean(stats::rnorm(40,mean=2,sd=1)))
stats 		Simulate the sampling distribution of a normal sample mean
Fixed workflow. mymean is a user-defined Stata program drawing 40 normal(2,1) observations and returning their mean; this is a fixed workflow.
Stata reference · R reference

Model reporting

Stata R Purpose, qualifications and sources 
estimates stats
 
broom::glance(fit)
broom 		Extract one-row model fit statistics
Workflow alternative returning an R table; it does not reproduce Stata table formatting.
Stata reference · R reference 
etable, cstat(_r_b) cstat(_r_se)
 
broom::tidy(fit,conf.int=TRUE)
broom 		Create a tidy coefficient table with uncertainty
Fixed workflow. Workflow alternative returning an R table; it does not reproduce Stata table formatting.
Stata reference · R reference 
predict result, residuals
 
broom::augment(fit)
broom 		Attach fitted values, residuals and influence diagnostics
Fixed workflow. Workflow alternative returning an R table; it does not reproduce Stata table formatting.
Stata reference · R reference

Spatial workflows

Stata R Purpose, qualifications and sources 
spbalance, balance
 
{ counts <- table(data$time); common <- as.numeric(names(counts)[counts == length(unique(data$id))]); data[data$time %in% common, , drop = FALSE] }
base 		Balance a spatial panel by retaining common time periods
Differs from Stata: Fixed workflow; adapt the named data columns explicitly. Requires unique unit/time pairs. Mirrors spbalance's choice to drop incomplete time periods while retaining spatial units; it does not fill missing outcomes.
Stata reference · R reference 
spdistance 1 2
 
sqrt(sum((as.numeric(data[data$id == 1, c("x", "y")]) - as.numeric(data[data$id == 2, c("x", "y")]))^2))
base 		Distance between two identified planar locations
Differs from Stata: Fixed workflow; adapt the named data columns explicitly. Planar Euclidean distance only; longitude/latitude require a geodesic calculation and explicit units.
Stata reference · R reference 
spgenerate nearby = W*value
 
drop(W %*% data$value)
base 		Compute a spatially lagged covariate
Differs from Stata: Fixed workflow; adapt the named data columns explicitly. W must use the identical unit ordering as data; this operation does not fit a spatial autoregressive model.
Stata reference · R reference 
spmatrix create idistance W, normalize(row)
 
{ D <- as.matrix(stats::dist(data[c("x", "y")])); W <- matrix(0, nrow(D), ncol(D)); off <- row(D) != col(D); W[off] <- 1 / D[off]; W / rowSums(W) }
stats 		Row-standardized inverse-distance spatial weights
Differs from Stata: Fixed workflow; adapt the named data columns explicitly. Planar distances, distinct non-coincident locations and matching row/column ID order are required. Stata defaults to spectral normalization, so row normalization is explicit.
Stata reference · R reference 
spset id, coord(x y)
 
sf::st_as_sf(data, coords = c("x", "y"), crs = 3857, remove = FALSE)
sf 		Declare planar point geometry and coordinate reference system
Differs from Stata: Fixed workflow; adapt the named data columns explicitly. Creating an sf object does not set Stata dataset metadata. Coordinates and CRS must describe the actual data; do not assign a projected CRS to longitude/latitude.
Stata reference · R reference

Causal inference

Stata R Purpose, qualifications and sources 
boottest treat, cluster(group) reps(199) weighttype(rademacher) seed(128)
 
fwildclusterboot::boottest(fit, param = "treat", clustid = "group", B = 199, type = "rademacher", impose_null = TRUE, conf_int = FALSE, engine = "R-lean")
fwildclusterboot 		Null-imposed wild cluster bootstrap using rademacher weights
Differs from Stata: Fixed workflow; adapt the named data columns explicitly. Small replication count is a runnable example, not an analysis recommendation. Align bootstrap variant, clustering and finite-sample correction. Installed fwildclusterboot emits a one-time seeding-change warning; both R and dqrng generators are explicitly seeded. fwildclusterboot 0.13 changed RNG seeding; seed-for-seed agreement with older versions is not expected.
Stata reference · R reference 
boottest treat, cluster(group) reps(199) weighttype(webb) seed(128)
 
fwildclusterboot::boottest(fit, param = "treat", clustid = "group", B = 199, type = "webb", impose_null = TRUE, conf_int = FALSE, engine = "R-lean")
fwildclusterboot 		Null-imposed wild cluster bootstrap using webb weights
Differs from Stata: Fixed workflow; adapt the named data columns explicitly. Small replication count is a runnable example, not an analysis recommendation. Align bootstrap variant, clustering and finite-sample correction. Installed fwildclusterboot emits a one-time seeding-change warning; both R and dqrng generators are explicitly seeded. fwildclusterboot 0.13 changed RNG seeding; seed-for-seed agreement with older versions is not expected.
Stata reference · R reference 
cem category, treatment(treat)
 
MatchIt::matchit(treat ~ category, data = transform(data, category = as.integer(category) - 1L), method = "cem", estimand = "ATT", cutpoints = list(category = 0))
MatchIt 		Coarsened exact matching on a categorical covariate
Fixed workflow; adapt the named data columns explicitly. Matching preprocessing only; categorical levels need no continuous cutpoints. For continuous variables, explicitly match cutpoints and estimand rather than relying on different software defaults.
Stata reference · R reference 
did_imputation y id time cohort
 
{ baseline <- stats::lm(y ~ factor(id) + factor(time) + x, data = data, subset = treated == 0); effects <- data$y - stats::predict(baseline, newdata = data); mean(effects[data$treated == 1]) }
stats 		Average imputation treatment effect over treated observations
Fixed workflow; adapt the named data columns explicitly. Point-estimation workflow only; untreated-outcome model, weighting and supported treated observations must be chosen explicitly.
Stata reference · R reference 
did_imputation y id time cohort, allhorizons
 
{ untreated <- data$treated == 0; baseline <- stats::lm(y ~ factor(id) + factor(time) + x, data = data, subset = untreated); effects <- data$y - stats::predict(baseline, newdata = data); keep <- data$treated == 1; stats::aggregate(effects[keep], list(horizon = data$time[keep] - data$cohort[keep]), mean) }
stats 		Untreated-outcome imputation effects by event time
Fixed workflow; adapt the named data columns explicitly. Point-estimation illustration with unit/time effects learned only from untreated observations. Never-treated Stata cohort should be missing, whereas this R fixture uses an out-of-range sentinel. Does not reproduce did_imputation's influence-function variance, autosample or pretrend tests.
Stata reference · R reference 
drdid y x, ivar(id) time(time) treatment(treated) dripw
 
{ ps <- stats::fitted(stats::glm(treated ~ x, data = data, family = stats::binomial())); m0 <- stats::predict(stats::lm(dy ~ x, data = data, subset = treated == 0), data); weighted.mean(data$dy - m0, data$treated) - weighted.mean(data$dy - m0, (1 - data$treated) * ps / (1 - ps)) }
stats 		Two-period panel doubly robust DiD ATT point estimate
Fixed workflow; adapt the named data columns explicitly. Wide-data point-estimation workflow for the dripw variant: logistic propensity plus control-group outcome-change OLS. Stata expects long data. No influence-function variance or improved drimp estimator.
Stata reference · R reference 
drdid y x, ivar(id) time(time) treatment(treated) reg
 
{ m0 <- stats::predict(stats::lm(dy ~ x, data = data, subset = treated == 0), data); mean((data$dy - m0)[data$treated == 1]) }
stats 		Panel outcome-regression DiD ATT point estimate
Fixed workflow; adapt the named data columns explicitly. Two-period wide-data regression-adjustment workflow; requires conditional parallel trends and no anticipation. No standard errors.
Stata reference · R reference 
drdid y x, ivar(id) time(time) treatment(treated) stdipw
 
{ ps <- stats::fitted(stats::glm(treated ~ x, data = data, family = stats::binomial())); mean(data$dy[data$treated == 1]) - weighted.mean(data$dy, (1 - data$treated) * ps / (1 - ps)) }
stats 		Stabilized inverse-probability-weighted DiD ATT
Fixed workflow; adapt the named data columns explicitly. Two-period wide-data point estimate; no influence-function variance and no repeated-cross-section implementation.
Stata reference · R reference 
ebalance treat x z, targets(1)
 
WeightIt::weightit(treat ~ x + z, data = data, method = "ebal", estimand = "ATT", moments = 1)
WeightIt 		Entropy balancing treated/control covariate moments through order 1
Fixed workflow; adapt the named data columns explicitly. Preprocessing weights targeting the treated population; normalization, higher moments and optimization tolerances must match. No outcome standard errors.
Stata reference · R reference 
ebalance treat x z, targets(2)
 
WeightIt::weightit(treat ~ x + z, data = data, method = "ebal", estimand = "ATT", moments = 2)
WeightIt 		Entropy balancing treated/control covariate moments through order 2
Fixed workflow; adapt the named data columns explicitly. Preprocessing weights targeting the treated population; normalization, higher moments and optimization tolerances must match. No outcome standard errors.
Stata reference · R reference 
estat endogenous
 
summary(AER::ivreg(y ~ x + w | z + w, data = data), diagnostics = TRUE)$diagnostics
AER 		IV weak-instrument and endogeneity diagnostics
Differs from Stata: Fixed workflow; adapt the named data columns explicitly. Wu-Hausman tests depend on model assumptions; exactly identified models have no overidentification test.
Stata reference · R reference 
estat firststage
 
summary(stats::lm(x ~ z + w, data = data))
stats 		First-stage instrument relevance regression
Fixed workflow; adapt the named data columns explicitly. This is the full first-stage regression; a joint excluded-instrument F test may require a nested comparison.
Stata reference · R reference 
eventstudyinteract y lead3 lead2 lag0 lag1 lag2, absorb(id time) cohort(cohort) control_cohort(never)
 
{ fit <- fixest::feols(y ~ sunab(cohort, time, ref.p = -1) | id + time, data = data, vcov = ~id, nthreads = 1); stats::aggregate(fit, agg = "ATT") }
fixest 		Aggregate cohort-interacted post-treatment effects
Fixed workflow; adapt the named data columns explicitly. ATT aggregation is an R postestimation step; match event-time/cohort weights rather than averaging coefficients unweighted.
Stata reference · R reference 
eventstudyinteract y lead3 lead2 lag0 lag1 lag2, absorb(id time) cohort(cohort) control_cohort(never) covariates(x) vce(cluster id)
 
fixest::feols(y ~ sunab(cohort, time, ref.p = -1) + x | id + time, data = data, vcov = ~id, nthreads = 1)
fixest 		Sun-Abraham cohort-interacted event-study estimation
Differs from Stata: Fixed workflow; adapt the named data columns explicitly. Uses never-treated reference cohorts and -1 event-time reference. Match supported horizons, cohort-share aggregation and covariance conventions to eventstudyinteract.
Stata reference · R reference 
gencohort cohort, group(id) time(time) treat(treated)
 
transform(data, cohort = ave(ifelse(treated == 1, time, Inf), id, FUN = function(v) { first <- min(v); rep(if(is.infinite(first)) 0 else first, length(v)) }))
stats 		Create first-treatment cohorts including never-treated units
Differs from Stata: Fixed workflow; adapt the named data columns explicitly. Uses zero as a never-treated category for cohort-based DiD. An absorbing treatment history and unique unit/time records are prerequisites; imputation estimators may instead require missing never-treatment times.
Stata reference · R reference 
ivreghdfe y z (x = instrument), absorb(id time)
 
fixest::feols(y ~ z | id + time | x ~ instrument, data = data, vcov = "iid", nthreads = 1)
fixest 		Instrumental variables with two absorbed effects
Differs from Stata: Fixed workflow; adapt the named data columns explicitly. User-written Stata command. Match singleton handling, covariance correction and absorbed-effect degrees of freedom before comparing inferential output.
Stata reference · R reference 
ivreghdfe y z (x = instrument), absorb(id time) cluster(id)
 
fixest::feols(y ~ z | id + time | x ~ instrument, data = data, vcov = ~id, nthreads = 1)
fixest 		Cluster-robust absorbed-effects IV
Differs from Stata: Fixed workflow; adapt the named data columns explicitly. User-written Stata command. Match singleton handling, covariance correction and absorbed-effect degrees of freedom before comparing inferential output.
Stata reference · R reference 
ivreghdfe y z (x = instrument), absorb(id time) robust
 
fixest::feols(y ~ z | id + time | x ~ instrument, data = data, vcov = "hetero", nthreads = 1)
fixest 		Heteroskedasticity-robust absorbed-effects IV
Differs from Stata: Fixed workflow; adapt the named data columns explicitly. User-written Stata command. Match singleton handling, covariance correction and absorbed-effect degrees of freedom before comparing inferential output.
Stata reference · R reference 
ivreghdfe y z (x = instrument), absorb(id)
 
fixest::feols(y ~ z | id | x ~ instrument, data = data, vcov = "iid", nthreads = 1)
fixest 		Instrumental variables with individual fixed effects
Differs from Stata: Fixed workflow; adapt the named data columns explicitly. User-written Stata command. Match singleton handling, covariance correction and absorbed-effect degrees of freedom before comparing inferential output.
Stata reference · R reference 
ivregress 2sls y w (x = z)
 
AER::ivreg(y ~ x + w | z + w, data = data)
AER 		Formula-based instrumental-variable regression
Differs from Stata: Fixed workflow; adapt the named data columns explicitly. Separate structural regressors and instruments with the formula bar.
Stata reference · R reference 
ivregress 2sls y w (x = z), vce(cluster cluster)
 
estimatr::iv_robust(y ~ x + w | z + w, data = data, clusters = cluster, se_type = "stata")
estimatr 		Cluster-robust two-stage least squares
Differs from Stata: Fixed workflow; adapt the named data columns explicitly. Small-cluster reference degrees of freedom are explicit.
Stata reference · R reference 
ivregress 2sls y w (x = z), vce(robust)
 
estimatr::iv_robust(y ~ x + w | z + w, data = data, se_type = "HC2")
estimatr 		Two-stage least squares with HC2 covariance
Differs from Stata: Fixed workflow; adapt the named data columns explicitly. HC corrections differ; estimatr stata aliases HC1 for unclustered IV inference.
Stata reference · R reference 
ivregress 2sls y w (x = z), vce(robust)
 
estimatr::iv_robust(y ~ x + w | z + w, data = data, se_type = "HC3")
estimatr 		Two-stage least squares with HC3 covariance
Differs from Stata: Fixed workflow; adapt the named data columns explicitly. HC corrections differ; estimatr stata aliases HC1 for unclustered IV inference.
Stata reference · R reference 
ivregress 2sls y w (x = z), vce(robust) small
 
estimatr::iv_robust(y ~ x + w | z + w, data = data, se_type = "HC1")
estimatr 		Two-stage least squares with HC1 covariance
Differs from Stata: Stata small requests the N/(N-K) HC1 correction and t-based inference. estimatr se_type="stata" aliases HC1 for unclustered IV. Without small, ivregress vce(robust) uses HC0 and normal reference inference, so it is a different default.
Stata reference · R reference 
ivregress 2sls y w (x = z), vce(robust) small
 
estimatr::iv_robust(y ~ x + w | z + w, data = data, se_type = "stata")
estimatr 		Two-stage least squares with stata covariance
Differs from Stata: Stata small requests the N/(N-K) HC1 correction and t-based inference. estimatr se_type="stata" aliases HC1 for unclustered IV. Without small, ivregress vce(robust) uses HC0 and normal reference inference, so it is a different default.
Stata reference · R reference 
kmatch eb treat x z (y), att targets(1)
 
WeightIt::weightit(treat ~ x + z, data = data, method = "ebal", estimand = "ATT", moments = 1)
WeightIt 		Entropy balancing treated/control covariate moments through order 1
Fixed workflow; adapt the named data columns explicitly. Preprocessing weights targeting the treated population; normalization, higher moments and optimization tolerances must match. No outcome standard errors.
Stata reference · R reference 
kmatch eb treat x z (y), att targets(2)
 
WeightIt::weightit(treat ~ x + z, data = data, method = "ebal", estimand = "ATT", moments = 2)
WeightIt 		Entropy balancing treated/control covariate moments through order 2
Fixed workflow; adapt the named data columns explicitly. Preprocessing weights targeting the treated population; normalization, higher moments and optimization tolerances must match. No outcome standard errors.
Stata reference · R reference 
kmatch em treat category (y), att
 
MatchIt::matchit(treat ~ category, data = data, method = "exact", estimand = "ATT")
MatchIt 		Exact matching on categorical covariates
Fixed workflow; adapt the named data columns explicitly. Exact-match preprocessing; estimand and stratum normalization are explicit.
Stata reference · R reference 
kmatch ipw treat x z (y), att
 
WeightIt::weightit(treat ~ x + z, data = data, method = "glm", estimand = "ATT")
WeightIt 		Inverse-probability weighting for the treated population
Fixed workflow; adapt the named data columns explicitly. Weights only; outcome contrasts and propensity-estimation variance require a separate estimator.
Stata reference · R reference 
kmatch md treat x z (y), att nn(1)
 
MatchIt::matchit(treat ~ x + z, data = data, method = "nearest", distance = "mahalanobis", replace = TRUE, estimand = "ATT")
MatchIt 		Multivariate Mahalanobis matching
Fixed workflow; adapt the named data columns explicitly. Explicit nearest-neighbor alternative to kmatch's kernel defaults; matching preprocessing only.
Stata reference · R reference 
kmatch ps treat x z (y), att nn(1)
 
MatchIt::matchit(treat ~ x + z, data = data, method = "nearest", distance = "glm", replace = TRUE, estimand = "ATT")
MatchIt 		Propensity-score nearest-neighbor matching
Fixed workflow; adapt the named data columns explicitly. Explicit nearest-neighbor alternative to kmatch's kernel defaults; matching preprocessing only.
Stata reference · R reference 
psmatch2 treat x z, outcome(y) logit neighbor(1)
 
MatchIt::matchit(treat ~ x + z, data = data, method = "nearest", distance = "glm", link = "logit", replace = TRUE, estimand = "ATT")
MatchIt 		Propensity-score nearest matching with replacement
Fixed workflow; adapt the named data columns explicitly. Returns matching preprocessing, not psmatch2's outcome estimate or variance. Replacement, tie-breaking and match order must be explicit.
Stata reference · R reference 
psmatch2 treat x z, outcome(y) logit neighbor(1)
 
{ match <- MatchIt::matchit(treat ~ x + z, data = data, method = "nearest", distance = "glm", replace = TRUE, estimand = "ATT"); matched <- MatchIt::match_data(match); with(matched, weighted.mean(y[treat == 1], weights[treat == 1]) - weighted.mean(y[treat == 0], weights[treat == 0])) }
MatchIt 		Matched ATT point contrast after propensity-score preprocessing
Fixed workflow; adapt the named data columns explicitly. Point contrast only; variance is not estimated, and matching ties/order may differ.
Stata reference · R reference 
psmatch2 treat x z, outcome(y) logit neighbor(1) caliper(.1)
 
MatchIt::matchit(treat ~ x + z, data = data, method = "nearest", distance = "glm", link = "logit", replace = TRUE, caliper = .1, std.caliper = FALSE, estimand = "ATT")
MatchIt 		Propensity-score matching with an absolute caliper
Fixed workflow; adapt the named data columns explicitly. Absolute probability-scale caliper; MatchIt defaults to standardized calipers unless explicitly disabled. Outcome estimation remains separate.
Stata reference · R reference 
psmatch2 treat x z, outcome(y) logit neighbor(1) noreplacement
 
MatchIt::matchit(treat ~ x + z, data = data, method = "nearest", distance = "glm", link = "logit", replace = FALSE, estimand = "ATT")
MatchIt 		Propensity-score nearest matching without replacement
Fixed workflow; adapt the named data columns explicitly. Returns matching preprocessing, not psmatch2's outcome estimate or variance. Replacement, tie-breaking and match order must be explicit.
Stata reference · R reference 
psmatch2 treat, outcome(y) mahalanobis(x z) neighbor(1)
 
MatchIt::matchit(treat ~ x + z, data = data, method = "nearest", distance = "mahalanobis", replace = TRUE, estimand = "ATT")
MatchIt 		Mahalanobis matching with replacement
Fixed workflow; adapt the named data columns explicitly. Matching preprocessing; covariance normalization and ties may differ from psmatch2.
Stata reference · R reference 
pstest x z, both
 
cobalt::bal.tab(match, un = TRUE)
cobalt 		Standardized covariate balance before and after matching
Fixed workflow; adapt the named data columns explicitly. Companion command supplied with psmatch2. Choose the same standardization denominator; cobalt's balance table does not reproduce every pstest t-test or omnibus summary.
Stata reference · R reference 
rdbwdensity x
 
rddensity::rdbwdensity(x)
rddensity 		Bandwidth selection for density discontinuity
Differs from Stata: Fixed workflow; adapt the named data columns explicitly. Bandwidths optimize density-estimation criteria, not treatment-effect estimation.
Stata reference · R reference 
rdbwselect y x, bwselect(cerrd)
 
rdrobust::rdbwselect(y, x, bwselect = "cerrd")
rdrobust 		RD bandwidth selection: cerrd
Differs from Stata: Fixed workflow; adapt the named data columns explicitly. Official rdrobust companion command; bandwidth selectors target different criteria.
Stata reference · R reference 
rdbwselect y x, bwselect(certwo)
 
rdrobust::rdbwselect(y, x, bwselect = "certwo")
rdrobust 		RD bandwidth selection: certwo
Differs from Stata: Fixed workflow; adapt the named data columns explicitly. Official rdrobust companion command; bandwidth selectors target different criteria.
Stata reference · R reference 
rdbwselect y x, bwselect(mserd)
 
rdrobust::rdbwselect(y, x, bwselect = "mserd")
rdrobust 		RD bandwidth selection: mserd
Differs from Stata: Fixed workflow; adapt the named data columns explicitly. Official rdrobust companion command; bandwidth selectors target different criteria.
Stata reference · R reference 
rdbwselect y x, bwselect(msetwo)
 
rdrobust::rdbwselect(y, x, bwselect = "msetwo")
rdrobust 		RD bandwidth selection: msetwo
Differs from Stata: Fixed workflow; adapt the named data columns explicitly. Official rdrobust companion command; bandwidth selectors target different criteria.
Stata reference · R reference 
rddensity x
 
rddensity::rddensity(x)
rddensity 		Density continuity at the treatment threshold
Differs from Stata: Fixed workflow; adapt the named data columns explicitly. Tests density continuity, not the RD identification assumptions as a whole.
Stata reference · R reference 
rddensity x, h(.4 .4)
 
rddensity::rddensity(x, h = c(.4, .4))
rddensity 		Density continuity with fixed bandwidths
Differs from Stata: Fixed workflow; adapt the named data columns explicitly. Tests density continuity, not the RD identification assumptions as a whole.
Stata reference · R reference 
rddensity x, p(3)
 
rddensity::rddensity(x, p = 3)
rddensity 		Cubic local-polynomial density discontinuity
Differs from Stata: Fixed workflow; adapt the named data columns explicitly. Tests density continuity, not the RD identification assumptions as a whole.
Stata reference · R reference 
rdrobust y x
 
rdrobust::rdrobust(y, x)
rdrobust 		Sharp local-linear RD
Differs from Stata: Fixed workflow; adapt the named data columns explicitly. User-written Stata rdrobust command. Match version, bandwidth, kernel and variance settings; no finite-fixture cross-software parity claim.
Stata reference · R reference 
rdrobust y x, bwselect(cerrd)
 
rdrobust::rdrobust(y, x, bwselect = "cerrd")
rdrobust 		Coverage-error-optimal RD bandwidth
Differs from Stata: Fixed workflow; adapt the named data columns explicitly. User-written Stata rdrobust command. Match version, bandwidth, kernel and variance settings; no finite-fixture cross-software parity claim.
Stata reference · R reference 
rdrobust y x, bwselect(msetwo)
 
rdrobust::rdrobust(y, x, bwselect = "msetwo")
rdrobust 		Separate MSE-optimal bandwidths by side
Differs from Stata: Fixed workflow; adapt the named data columns explicitly. User-written Stata rdrobust command. Match version, bandwidth, kernel and variance settings; no finite-fixture cross-software parity claim.
Stata reference · R reference 
rdrobust y x, cluster(cluster)
 
rdrobust::rdrobust(y, x, cluster = cluster)
rdrobust 		Cluster-robust RD inference
Differs from Stata: Fixed workflow; adapt the named data columns explicitly. User-written Stata rdrobust command. Match version, bandwidth, kernel and variance settings; no finite-fixture cross-software parity claim.
Stata reference · R reference 
rdrobust y x, covs(z)
 
rdrobust::rdrobust(y, x, covs = z)
rdrobust 		Covariate-adjusted RD
Differs from Stata: Fixed workflow; adapt the named data columns explicitly. User-written Stata rdrobust command. Match version, bandwidth, kernel and variance settings; no finite-fixture cross-software parity claim.
Stata reference · R reference 
rdrobust y x, fuzzy(treat)
 
rdrobust::rdrobust(y, x, fuzzy = treat)
rdrobust 		Fuzzy regression discontinuity
Differs from Stata: Fixed workflow; adapt the named data columns explicitly. Requires an exclusion restriction, a treatment-probability discontinuity and a suitable monotonicity assumption.
Stata reference · R reference 
rdrobust y x, h(.4)
 
rdrobust::rdrobust(y, x, h = .4)
rdrobust 		RD with fixed symmetric bandwidth
Differs from Stata: Fixed workflow; adapt the named data columns explicitly. User-written Stata rdrobust command. Match version, bandwidth, kernel and variance settings; no finite-fixture cross-software parity claim.
Stata reference · R reference 
rdrobust y x, kernel(epanechnikov)
 
rdrobust::rdrobust(y, x, kernel = "epanechnikov")
rdrobust 		RD with Epanechnikov kernel
Differs from Stata: Fixed workflow; adapt the named data columns explicitly. User-written Stata rdrobust command. Match version, bandwidth, kernel and variance settings; no finite-fixture cross-software parity claim.
Stata reference · R reference 
rdrobust y x, kernel(uniform)
 
rdrobust::rdrobust(y, x, kernel = "uniform")
rdrobust 		RD with uniform kernel
Differs from Stata: Fixed workflow; adapt the named data columns explicitly. User-written Stata rdrobust command. Match version, bandwidth, kernel and variance settings; no finite-fixture cross-software parity claim.
Stata reference · R reference 
rdrobust y x, level(90)
 
rdrobust::rdrobust(y, x, level = 90)
rdrobust 		RD with ninety-percent confidence intervals
Differs from Stata: Fixed workflow; adapt the named data columns explicitly. User-written Stata rdrobust command. Match version, bandwidth, kernel and variance settings; no finite-fixture cross-software parity claim.
Stata reference · R reference 
rdrobust y x, p(2)
 
rdrobust::rdrobust(y, x, p = 2)
rdrobust 		Local-quadratic RD
Differs from Stata: Fixed workflow; adapt the named data columns explicitly. User-written Stata rdrobust command. Match version, bandwidth, kernel and variance settings; no finite-fixture cross-software parity claim.
Stata reference · R reference 
tebalance summarize
 
cobalt::bal.tab(w, un = TRUE)
cobalt 		Covariate balance before and after weighting
Differs from Stata: Fixed workflow; adapt the named data columns explicitly. Standardization denominator and binary-variable scaling must be aligned.
Stata reference · R reference 
teffects aipw (y x z) (treat x z)
 
{ ps <- stats::fitted(stats::glm(treat ~ x + z, data, family = stats::binomial())); m0 <- stats::predict(stats::lm(y ~ x + z, data, subset = treat == 0), data); m1 <- stats::predict(stats::lm(y ~ x + z, data, subset = treat == 1), data); mean(m1 - m0 + data$treat * (data$y - m1) / ps - (1 - data$treat) * (data$y - m0) / (1 - ps)) }
stats 		Augmented inverse-probability ATE point estimate
Differs from Stata: Fixed workflow; adapt the named data columns explicitly. No standard errors; double robustness still requires overlap and identifying assumptions.
Stata reference · R reference 
teffects ipw (y) (treat x z)
 
WeightIt::weightit(treat ~ x + z, data = data, method = "glm", estimand = "ATE")
WeightIt 		Propensity-score weights targeting ATE
Fixed workflow; adapt the named data columns explicitly. Computes weights only, not treatment effects or standard errors. ATO and ATC are alternative target populations, not Stata teffects defaults.
Stata reference · R reference 
teffects ipw (y) (treat x z)
 
WeightIt::weightit(treat ~ x + z, data = data, method = "glm", estimand = "ATC")
WeightIt 		Propensity-score weights targeting ATC
Fixed workflow; adapt the named data columns explicitly. Computes weights only, not treatment effects or standard errors. ATO and ATC are alternative target populations, not Stata teffects defaults.
Stata reference · R reference 
teffects ipw (y) (treat x z)
 
WeightIt::weightit(treat ~ x + z, data = data, method = "glm", estimand = "ATO")
WeightIt 		Propensity-score weights targeting ATO
Fixed workflow; adapt the named data columns explicitly. Computes weights only, not treatment effects or standard errors. ATO and ATC are alternative target populations, not Stata teffects defaults.
Stata reference · R reference 
teffects ipw (y) (treat x z)
 
{ w <- WeightIt::weightit(treat ~ x + z, data = data, method = "glm", estimand = "ATE")$weights; weighted.mean(data$y[data$treat == 1], w[data$treat == 1]) - weighted.mean(data$y[data$treat == 0], w[data$treat == 0]) }
WeightIt 		Normalized IPW point effect for ATE
Differs from Stata: Fixed workflow; adapt the named data columns explicitly. Point estimate only; outcome-free weighting assumes conditional exchangeability and overlap. Weight-estimation uncertainty is not quantified.
Stata reference · R reference 
teffects ipw (y) (treat x z) , atet
 
WeightIt::weightit(treat ~ x + z, data = data, method = "glm", estimand = "ATT")
WeightIt 		Propensity-score weights targeting ATT
Fixed workflow; adapt the named data columns explicitly. Computes weights only, not treatment effects or standard errors. ATO and ATC are alternative target populations, not Stata teffects defaults.
Stata reference · R reference 
teffects ipw (y) (treat x z) , atet
 
{ w <- WeightIt::weightit(treat ~ x + z, data = data, method = "glm", estimand = "ATT")$weights; weighted.mean(data$y[data$treat == 1], w[data$treat == 1]) - weighted.mean(data$y[data$treat == 0], w[data$treat == 0]) }
WeightIt 		Normalized IPW point effect for ATT
Differs from Stata: Fixed workflow; adapt the named data columns explicitly. Point estimate only; outcome-free weighting assumes conditional exchangeability and overlap. Weight-estimation uncertainty is not quantified.
Stata reference · R reference 
teffects ipwra (y x z) (treat x z)
 
{ ps <- stats::fitted(stats::glm(treat ~ x + z, data, family = stats::binomial())); data$w <- ifelse(data$treat == 1, 1 / ps, 1 / (1 - ps)); m0 <- stats::predict(stats::lm(y ~ x + z, data, subset = treat == 0, weights = w), data); m1 <- stats::predict(stats::lm(y ~ x + z, data, subset = treat == 1, weights = w), data); mean(m1 - m0) }
stats 		Inverse-probability-weighted outcome-regression ATE
Differs from Stata: Fixed workflow; adapt the named data columns explicitly. Point estimation only; Stata jointly computes the estimating-equation variance.
Stata reference · R reference 
teffects nnmatch (y x z) (treat), atet
 
MatchIt::matchit(treat ~ x + z, data = data, method = "nearest", distance = "mahalanobis", replace = TRUE, estimand = "ATT")
MatchIt 		Mahalanobis nearest-neighbor preprocessing
Matches with replacement, as teffects does. MatchIt returns preprocessing weights and matches, not the treatment-effect estimate; tie handling, distance conventions, bias correction and Abadie-Imbens standard errors require separate alignment.
Stata reference · R reference 
teffects psmatch (y) (treat x z), nneighbor(1) atet
 
MatchIt::matchit(treat ~ x + z, data = data, method = "nearest", distance = "glm", ratio = 1, replace = TRUE, estimand = "ATT")
MatchIt 		One-to-one nearest propensity-score matching
Matches with replacement, as teffects does. MatchIt returns preprocessing weights and matches, not the treatment-effect estimate; tie handling, distance conventions, bias correction and Abadie-Imbens standard errors require separate alignment.
Stata reference · R reference 
teffects psmatch (y) (treat x z), nneighbor(1) atet
 
MatchIt::matchit(treat ~ x + z, data = data, method = "nearest", distance = "glm", ratio = 1, replace = TRUE, estimand = "ATT")
MatchIt 		Matching with replacement
Fixed workflow; adapt the named data columns explicitly. Preprocessing only; MatchIt does not reproduce teffects matching bias correction or Abadie-Imbens standard errors.
Stata reference · R reference 
teffects psmatch (y) (treat x z), nneighbor(2) atet
 
MatchIt::matchit(treat ~ x + z, data = data, method = "nearest", distance = "glm", ratio = 2, replace = TRUE, estimand = "ATT")
MatchIt 		One-to-two nearest propensity-score matching
Matches with replacement, as teffects does. MatchIt returns preprocessing weights and matches, not the treatment-effect estimate; tie handling, distance conventions, bias correction and Abadie-Imbens standard errors require separate alignment.
Stata reference · R reference 
teffects ra (y x z) (treat)
 
{ fit0 <- stats::lm(y ~ x + z, data, subset = treat == 0); fit1 <- stats::lm(y ~ x + z, data, subset = treat == 1); mean(stats::predict(fit1, data) - stats::predict(fit0, data)) }
stats 		Regression-adjustment average treatment effect
Differs from Stata: Fixed workflow; adapt the named data columns explicitly. Point-estimate g-computation under correctly specified potential-outcome models; no causal identification is established by fitting alone.
Stata reference · R reference 
teffects ra (y x z) (treat), atet
 
{ fit0 <- stats::lm(y ~ x + z, data, subset = treat == 0); fit1 <- stats::lm(y ~ x + z, data, subset = treat == 1); treated <- data[data$treat == 1, ]; mean(stats::predict(fit1, treated) - stats::predict(fit0, treated)) }
stats 		Regression-adjustment effect among treated units
Differs from Stata: Fixed workflow; adapt the named data columns explicitly. Averages counterfactual differences over the treated covariate distribution.
Stata reference · R reference 
teoverlap
 
{ ps <- stats::fitted(stats::glm(treat ~ x + z, data, family = stats::binomial())); lapply(split(ps, data$treat), function(v) stats::quantile(v, c(0, .05, .5, .95, 1))) }
stats 		Propensity-score overlap summary by treatment
Fixed workflow; adapt the named data columns explicitly. Numeric overlap diagnostic; no density plot or automatic trimming.
Stata reference · R reference 
xtdidregress (y x) (treated), group(id) time(time)
 
fixest::feols(y ~ treated + x | id + time, data = data, vcov = ~id, nthreads = 1)
fixest 		Panel difference-in-differences with unit and period fixed effects
Differs from Stata: Fixed workflow; adapt the named data columns explicitly. Homogeneous-effect two-group design with parallel trends and no anticipation. Under staggered timing and heterogeneous effects, TWFE is not a replacement for modern cohort-specific estimators; covariance finite-sample conventions can differ.
Stata reference · R reference

Panel data

Stata R Purpose, qualifications and sources 
hausman fixed random
 
plm::phtest(fe, re)
plm 		Hausman comparison of fixed and random effects
Differs from Stata: Fixed workflow; adapt the named data columns explicitly. Requires same regressors and compatible covariance assumptions.
Stata reference · R reference 
xtcloglog binary x, re
 
lme4::glmer(binary ~ x + (1 | id), data = data, family = stats::binomial(link = "cloglog"), nAGQ = 12)
lme4 		Random-intercept panel cloglog
Differs from Stata: Uses normal panel random intercepts and 12-point adaptive quadrature, matching Stata intpoints(12). Quadrature adaptation and convergence criteria remain package-specific.
Stata reference · R reference 
xtdpd y L.y x, div(x) dgmmiv(y, lagrange(2 3))
 
{ plm <- plm::plm; plm::pgmm(y ~ lag(y, 1) + x | lag(y, 2:3), data = pd, effect = "individual", model = "onestep", transformation = "d", collapse = TRUE) }
plm 		One-step difference GMM for a dynamic panel
Differs from Stata: Fixed workflow; adapt the named data columns explicitly. Supports the standard difference-GMM subset, with deliberately collapsed instruments. Stata xtdpd supports richer error processes; align instruments, initial weight matrices and finite-sample corrections before comparing.
Stata reference · R reference 
xtdpdsys y x, maxldep(3)
 
{ plm <- plm::plm; plm::pgmm(y ~ lag(y, 1) + x | lag(y, 2:3), data = pd, effect = "individual", model = "onestep", transformation = "ld", collapse = TRUE) }
plm 		System GMM with level and differenced equations
Differs from Stata: Fixed workflow; adapt the named data columns explicitly. Additional level moment conditions require stronger initial-condition assumptions. Instrument collapsing, instrument ranges and weighting differ unless explicitly matched; no blanket parity with xtdpdsys defaults.
Stata reference · R reference 
xtgee binary x, family( binomial ) corr( ar1 )
 
geepack::geeglm(binary ~ x, id = id, waves = time, data = data, family = stats::binomial(), corstr = "ar1")
geepack 		binomial GEE with ar1 working correlation
Differs from Stata: Keep rows sorted in contiguous panels. geeglm defaults to sandwich robust standard errors (std.err="san.se"), corresponding to xtgee, vce(robust); Stata defaults to conventional model-based standard errors. Scale estimation, small-sample adjustments and correlation estimation may still differ.
Stata reference · R reference 
xtgee binary x, family( binomial ) corr( exchangeable )
 
geepack::geeglm(binary ~ x, id = id, waves = time, data = data, family = stats::binomial(), corstr = "exchangeable")
geepack 		binomial GEE with exchangeable working correlation
Differs from Stata: Keep rows sorted in contiguous panels. geeglm defaults to sandwich robust standard errors (std.err="san.se"), corresponding to xtgee, vce(robust); Stata defaults to conventional model-based standard errors. Scale estimation, small-sample adjustments and correlation estimation may still differ.
Stata reference · R reference 
xtgee binary x, family( binomial ) corr( independence )
 
geepack::geeglm(binary ~ x, id = id, waves = time, data = data, family = stats::binomial(), corstr = "independence")
geepack 		binomial GEE with independence working correlation
Differs from Stata: Keep rows sorted in contiguous panels. geeglm defaults to sandwich robust standard errors (std.err="san.se"), corresponding to xtgee, vce(robust); Stata defaults to conventional model-based standard errors. Scale estimation, small-sample adjustments and correlation estimation may still differ.
Stata reference · R reference 
xtgee count x, family( poisson ) corr( ar1 )
 
geepack::geeglm(count ~ x, id = id, waves = time, data = data, family = stats::poisson(), corstr = "ar1")
geepack 		poisson GEE with ar1 working correlation
Differs from Stata: Keep rows sorted in contiguous panels. geeglm defaults to sandwich robust standard errors (std.err="san.se"), corresponding to xtgee, vce(robust); Stata defaults to conventional model-based standard errors. Scale estimation, small-sample adjustments and correlation estimation may still differ.
Stata reference · R reference 
xtgee count x, family( poisson ) corr( exchangeable )
 
geepack::geeglm(count ~ x, id = id, waves = time, data = data, family = stats::poisson(), corstr = "exchangeable")
geepack 		poisson GEE with exchangeable working correlation
Differs from Stata: Keep rows sorted in contiguous panels. geeglm defaults to sandwich robust standard errors (std.err="san.se"), corresponding to xtgee, vce(robust); Stata defaults to conventional model-based standard errors. Scale estimation, small-sample adjustments and correlation estimation may still differ.
Stata reference · R reference 
xtgee count x, family( poisson ) corr( independence )
 
geepack::geeglm(count ~ x, id = id, waves = time, data = data, family = stats::poisson(), corstr = "independence")
geepack 		poisson GEE with independence working correlation
Differs from Stata: Keep rows sorted in contiguous panels. geeglm defaults to sandwich robust standard errors (std.err="san.se"), corresponding to xtgee, vce(robust); Stata defaults to conventional model-based standard errors. Scale estimation, small-sample adjustments and correlation estimation may still differ.
Stata reference · R reference 
xtgee y x, family( gaussian ) corr( ar1 )
 
geepack::geeglm(y ~ x, id = id, waves = time, data = data, family = stats::gaussian(), corstr = "ar1")
geepack 		gaussian GEE with ar1 working correlation
Differs from Stata: Keep rows sorted in contiguous panels. geeglm defaults to sandwich robust standard errors (std.err="san.se"), corresponding to xtgee, vce(robust); Stata defaults to conventional model-based standard errors. Scale estimation, small-sample adjustments and correlation estimation may still differ.
Stata reference · R reference 
xtgee y x, family( gaussian ) corr( exchangeable )
 
geepack::geeglm(y ~ x, id = id, waves = time, data = data, family = stats::gaussian(), corstr = "exchangeable")
geepack 		gaussian GEE with exchangeable working correlation
Differs from Stata: Keep rows sorted in contiguous panels. geeglm defaults to sandwich robust standard errors (std.err="san.se"), corresponding to xtgee, vce(robust); Stata defaults to conventional model-based standard errors. Scale estimation, small-sample adjustments and correlation estimation may still differ.
Stata reference · R reference 
xtgee y x, family( gaussian ) corr( independence )
 
geepack::geeglm(y ~ x, id = id, waves = time, data = data, family = stats::gaussian(), corstr = "independence")
geepack 		gaussian GEE with independence working correlation
Differs from Stata: Keep rows sorted in contiguous panels. geeglm defaults to sandwich robust standard errors (std.err="san.se"), corresponding to xtgee, vce(robust); Stata defaults to conventional model-based standard errors. Scale estimation, small-sample adjustments and correlation estimation may still differ.
Stata reference · R reference 
xtgls y x z, panels(heteroskedastic) igls
 
nlme::gls(y ~ x + z, data = data, weights = nlme::varIdent(form = ~1 | id), method = "ML")
nlme 		Group-heteroskedastic panel GLS by maximum likelihood
Fixed workflow; adapt the named data columns explicitly. Common slopes with independent residuals and group-specific residual variances; likelihood, variance normalization and convergence criteria can differ.
Stata reference · R reference 
xthtaylor y x1 x2 a b, endog(x2 b) constant(a b)
 
plm::plm(y ~ x1 + x2 + a + b | x1 + a | x2, data = data, index = c("id", "time"), random.method = "ht", model = "random", inst.method = "baltagi")
plm 		Hausman-Taylor model with time-varying and invariant endogenous covariates
Differs from Stata: Fixed workflow; adapt the named data columns explicitly. The three-part R formula explicitly separates exogenous covariates and time-varying endogenous covariates. Identification of invariant endogenous effects requires sufficient exogenous time-varying instruments; these classifications need substantive justification.
Stata reference · R reference 
xtivreg y z (x = instrument), fe
 
plm::plm(y ~ x + z | instrument + z, data = pd, model = "within")
plm 		Fixed-effects instrumental-variable regression
Differs from Stata: Fixed workflow; adapt the named data columns explicitly. Instrument validity must be justified substantively; this fixture only verifies execution.
Stata reference · R reference 
xtlogit binary x, fe
 
{ coxph <- survival::coxph; survival::clogit(binary ~ x + survival::strata(id), data = data) }
survival 		Conditional fixed-effects logistic regression
Differs from Stata: Fixed workflow; adapt the named data columns explicitly. Conditions out group intercepts; groups without outcome variation contribute no coefficient information.
Stata reference · R reference 
xtpcse y x z, correlation(independent)
 
{ fit <- plm::plm(y ~ x + z, data = pd, model = "pooling"); lmtest::coeftest(fit, vcov. = plm::vcovBK(fit, type = "HC0", cluster = "time")) }
plm 		Beck-Katz panel-corrected covariance
Differs from Stata: vcovBK(cluster="time") allows contemporaneous cross-panel covariance, as required for PCSE. Balancedness, pairwise missing-data treatment, finite-sample normalization and reference tests can still differ from xtpcse.
Stata reference · R reference 
xtpoisson count x, re normal
 
lme4::glmer(count ~ x + (1 | id), data = data, family = stats::poisson(link = "log"), nAGQ = 12)
lme4 		Random-intercept panel Poisson
Differs from Stata: Uses normal panel random intercepts and 12-point adaptive quadrature, matching the Stata integration-point default. xtpoisson requires re normal for this distribution; its plain re default is gamma. Quadrature adaptation and convergence criteria remain package-specific.
Stata reference · R reference 
xtprobit binary x, re
 
lme4::glmer(binary ~ x + (1 | id), data = data, family = stats::binomial(link = "probit"), nAGQ = 12)
lme4 		Random-intercept panel probit
Differs from Stata: Uses normal panel random intercepts and 12-point adaptive quadrature, matching Stata intpoints(12). Quadrature adaptation and convergence criteria remain package-specific.
Stata reference · R reference 
xtrc y x z
 
nlme::lme(y ~ x + z, random = ~ x + z | id, data = data, method = "ML")
nlme 		Random-coefficient panel model workflow
All three coefficients (intercept, x and z) vary by panel. nlme fits a Gaussian mixed model by ML; xtrc uses Swamy random-coefficient GLS, so the estimators and covariance estimates differ.
Stata reference · R reference 
xtreg y x z i.time, fe
 
plm::plm(y ~ x + z, data = pd, model = "within", effect = "twoways")
plm 		Individual and time fixed effects
Differs from Stata: Fixed workflow; adapt the named data columns explicitly. First-difference and pooled rows are explicit benchmarks; match estimator and effects.
Stata reference · R reference 
xtreg y x z, be
 
plm::plm(y ~ x + z, data = pd, model = "between", effect = "individual")
plm 		Between-unit regression
Differs from Stata: Fixed workflow; adapt the named data columns explicitly. First-difference and pooled rows are explicit benchmarks; match estimator and effects.
Stata reference · R reference 
xtreg y x z, fe
 
plm::plm(y ~ x + z, data = pd, model = "within", effect = "individual")
plm 		Individual fixed effects
Differs from Stata: Fixed workflow; adapt the named data columns explicitly. First-difference and pooled rows are explicit benchmarks; match estimator and effects.
Stata reference · R reference 
xtreg y x z, fe
 
plm::plm(y ~ x + z, data = pd, model = "within", effect = "individual")
plm 		Fit a within-panel fixed-effects regression
Both commands fit individual fixed effects using the within transformation. plm reports the transformed-model coefficient vector without Stata reported intercept; residual degrees of freedom and covariance scaling must be aligned separately.
Stata reference · R reference 
xtreg y x z, fe vce(cluster id)
 
{ V <- plm::vcovHC(fit, method = "arellano", type = "HC1", cluster = "group"); lmtest::coeftest(fit, vcov. = V) }
plm 		Cluster-robust panel standard errors
Differs from Stata: Stata xtreg, fe vce(robust) equals vce(cluster id). plm HC1 uses N/(N-K) with K counting within-model coefficients and coeftest uses residual degrees of freedom; Stata also applies its cluster correction and uses G-1 reference degrees of freedom. These standard errors and p-values are not exact Stata finite-sample parity.
Stata reference · R reference 
xtreg y x z, fe vce(robust)
 
{ V <- plm::vcovHC(fit, method = "arellano", type = "HC1", cluster = "group"); lmtest::coeftest(fit, vcov. = V) }
plm 		Cluster-robust within-panel standard errors
Differs from Stata: Stata xtreg, fe vce(robust) equals vce(cluster id). plm HC1 uses N/(N-K) with K counting within-model coefficients and coeftest uses residual degrees of freedom; Stata also applies its cluster correction and uses G-1 reference degrees of freedom. These standard errors and p-values are not exact Stata finite-sample parity.
Stata reference · R reference 
xtreg y x z, pa corr(independent)
 
plm::plm(y ~ x + z, data = pd, model = "pooling", effect = "individual")
plm 		Pooled panel regression benchmark
Under independent working correlation, population-averaged Gaussian GEE and pooled OLS have the same coefficients. xtreg, pa defaults to conventional model-based GEE covariance (not robust unless requested); pooled plm uses OLS residual scaling, so reported standard errors can differ.
Stata reference · R reference 
xtreg y x z, re
 
plm::plm(y ~ x + z, data = pd, model = "random", effect = "individual")
plm 		Random-effects panel regression
Differs from Stata: Fixed workflow; adapt the named data columns explicitly. First-difference and pooled rows are explicit benchmarks; match estimator and effects.
Stata reference · R reference 
xtreg y x z, re
 
plm::plm(y ~ x + z, data = pd, model = "random", random.method = "swar")
plm 		Random effects with swar variance components
Fixed workflow; adapt the named data columns explicitly. These are named alternative variance-component estimators; only matching methods justify comparing estimates.
Stata reference · R reference 
xtreg y x z, re
 
plm::plm(y ~ x + z, data = pd, model = "random", random.method = "walhus")
plm 		Random effects with walhus variance components
Walhus is an R alternative variance-component estimator with no corresponding xtreg option. Stata xtreg, re uses its default estimator or sa for Swamy-Arora; the Stata line does not select Walhus and numerical parity is not claimed.
Stata reference · R reference 
xtreg y x z, re
 
plm::plm(y ~ x + z, data = pd, model = "random", random.method = "amemiya")
plm 		Random effects with amemiya variance components
Amemiya is an R alternative variance-component estimator with no corresponding xtreg option. Stata xtreg, re uses its default estimator or sa for Swamy-Arora; the Stata line does not select Amemiya and numerical parity is not claimed.
Stata reference · R reference 
xtreg y x z, re
 
plm::plm(y ~ x + z, data = pd, model = "random", random.method = "nerlove")
plm 		Random effects with nerlove variance components
Nerlove is an R alternative variance-component estimator with no corresponding xtreg option. Stata xtreg, re uses its default estimator or sa for Swamy-Arora; the Stata line does not select Nerlove and numerical parity is not claimed.
Stata reference · R reference 
xtregar y x z, re
 
nlme::lme(y ~ x + z, random = ~1 | id, correlation = nlme::corAR1(form = ~time | id), data = data, method = "ML")
nlme 		Random intercept with AR(1) residual correlation
Fixed workflow; adapt the named data columns explicitly. Likelihood-based mixed-model workflow differs from xtregar's estimator.
Stata reference · R reference 
xtscc y x z, fe lag(2)
 
{ V <- plm::vcovSCC(fit, type = "HC1", maxlag = 2); lmtest::coeftest(fit, vcov. = V) }
plm 		Driscoll-Kraay covariance
Differs from Stata: Requires the community command xtscc (ssc install xtscc); xtreg has no vce(dkraay) option. Both use a within estimator with lag-2 Driscoll-Kraay covariance; finite-sample scaling and reference degrees of freedom differ.
Stata reference · R reference 
xtreg y x z, re
xttest0
 
plm::plmtest(plm::plm(y ~ x + z, data = pd, model = "pooling"), type = "bp", effect = "individual")
plm 		Breusch-Pagan test of individual random effects
Differs from Stata: Fixed workflow; adapt the named data columns explicitly. Stata's boundary-adjusted mixture distribution can differ from plm's chi-square reference.
Stata reference · R reference 
xtunitroot fisher y, dfuller lags(1)
 
plm::purtest(pd$y, test = "madwu", exo = "intercept", lags = 1, pmax = 1)
plm 		Panel unit-root test: madwu
Differs from Stata: Fixed workflow; adapt the named data columns explicitly. Hadri reverses the null to stationarity; deterministic terms and lag rules must match.
Stata reference · R reference 
xtunitroot hadri y, robust
 
plm::purtest(pd$y, test = "hadri", exo = "intercept", Hcons = TRUE)
plm 		Panel unit-root test: hadri
Differs from Stata: The null is panel stationarity. Hcons=TRUE selects the heteroskedasticity-robust Hadri statistic, corresponding to Stata robust; plm ignores lag selections for Hadri and provides no Bartlett HAC kernel correction.
Stata reference · R reference 
xtunitroot ips y, lags(1)
 
plm::purtest(pd$y, test = "ips", exo = "intercept", lags = 1, pmax = 1)
plm 		Panel unit-root test: ips
Differs from Stata: Fixed workflow; adapt the named data columns explicitly. Hadri reverses the null to stationarity; deterministic terms and lag rules must match.
Stata reference · R reference 
xtunitroot llc y, lags(1)
 
plm::purtest(pd$y, test = "levinlin", exo = "intercept", lags = 1, pmax = 1)
plm 		Panel unit-root test: levinlin
Differs from Stata: Fixed workflow; adapt the named data columns explicitly. Hadri reverses the null to stationarity; deterministic terms and lag rules must match.
Stata reference · R reference

Time series and forecasting

Stata R Purpose, qualifications and sources 
ac y, lags(12)
 
stats::acf(y, lag.max = 12, plot = FALSE)
stats 		Autocorrelation function
Differs from Stata: Fixed workflow; adapt the named data columns explicitly. R returns numeric correlations; request plotting separately.
Stata reference · R reference 
arch y, arch(1) garch(1) noconstant
 
tseries::garch(as.numeric(y), order = c(1, 1), trace = FALSE)
tseries 		GARCH(1,1) conditional-variance model
Differs from Stata: Fixed workflow; adapt the named data columns explicitly. Fit a mean model separately; optimization and initialization differ.
Stata reference · R reference 
arch y, arch(1) noconstant
 
tseries::garch(as.numeric(y), order = c(0, 1), trace = FALSE)
tseries 		ARCH(1) conditional-variance model
Differs from Stata: Fixed workflow; adapt the named data columns explicitly. tseries::garch models zero-mean observations; remove or explicitly model the conditional mean first.
Stata reference · R reference 
arfima y, ar(1) ma(1)
 
fracdiff::fracdiff(y, nar = 1, nma = 1)
fracdiff 		Fractionally integrated ARMA estimation
Differs from Stata: Fixed workflow; adapt the named data columns explicitly. fracdiff uses approximate likelihood and a different MA sign convention.
Stata reference · R reference 
arfimasoc y, maxar(1) maxma(1)
 
{ grid <- expand.grid(p = 0:1, q = 0:1); grid$AIC <- mapply(function(p, q) { f <- fracdiff::fracdiff(y, nar = p, nma = q); -2 * f$log.likelihood + 2 * (p + q + 2) }, grid$p, grid$q); grid }
fracdiff 		Fractional-ARMA order comparison
Fixed workflow; adapt the named data columns explicitly. Approximate-likelihood order-selection workflow; no numerical parity claim.
Stata reference · R reference 
arima y x, arima(1,0,0)
 
stats::arima(y, order = c(1, 0, 0), xreg = x)
stats 		ARMAX with external regressor
Differs from Stata: Fixed workflow; adapt the named data columns explicitly. Parameterization, initialization and likelihood conventions differ; compare selected model, not default output.
Stata reference · R reference 
arima y, arima(0,0,0) sarima(0,1,1,4)
 
stats::arima(y, order = c(0, 0, 0), seasonal = list(order = c(0, 1, 1), period = 4))
stats 		Seasonally differenced moving average
Differs from Stata: Fixed workflow; adapt the named data columns explicitly. Parameterization, initialization and likelihood conventions differ; compare selected model, not default output.
Stata reference · R reference 
arima y, arima(0,0,0) sarima(1,0,0,4)
 
stats::arima(y, order = c(0, 0, 0), seasonal = list(order = c(1, 0, 0), period = 4))
stats 		Seasonal autoregression
Differs from Stata: Fixed workflow; adapt the named data columns explicitly. Parameterization, initialization and likelihood conventions differ; compare selected model, not default output.
Stata reference · R reference 
arima y, arima(0,0,1)
 
stats::arima(y, order = c(0, 0, 1))
stats 		MA(1) with intercept
Differs from Stata: Fixed workflow; adapt the named data columns explicitly. Parameterization, initialization and likelihood conventions differ; compare selected model, not default output.
Stata reference · R reference 
arima y, arima(0,1,1)
 
stats::arima(y, order = c(0, 1, 1))
stats 		Integrated moving-average model
Differs from Stata: Fixed workflow; adapt the named data columns explicitly. Parameterization, initialization and likelihood conventions differ; compare selected model, not default output.
Stata reference · R reference 
arima y, arima(1,0,0)
 
stats::arima(y, order = c(1, 0, 0))
stats 		AR(1) with intercept
Differs from Stata: Fixed workflow; adapt the named data columns explicitly. Parameterization, initialization and likelihood conventions differ; compare selected model, not default output.
Stata reference · R reference 
arima y, arima(1,0,0) diffuse
 
stats::arima(y, order = c(1, 0, 0), method = "ML", SSinit = "Rossignol2011")
stats 		State-space maximum-likelihood estimation
Differs from Stata: Fixed workflow; adapt the named data columns explicitly. Parameterization, initialization and likelihood conventions differ; compare selected model, not default output.
Stata reference · R reference 
arima y, arima(1,0,0) noconstant
 
stats::arima(y, order = c(1, 0, 0), include.mean = FALSE)
stats 		Zero-mean autoregression
Differs from Stata: Fixed workflow; adapt the named data columns explicitly. Parameterization, initialization and likelihood conventions differ; compare selected model, not default output.
Stata reference · R reference 
arima y, arima(1,0,1)
 
stats::arima(y, order = c(1, 0, 1))
stats 		ARMA(1,1)
Differs from Stata: Fixed workflow; adapt the named data columns explicitly. Parameterization, initialization and likelihood conventions differ; compare selected model, not default output.
Stata reference · R reference 
arima y, arima(1,0,1)
 
stats::arima(y, order = c(1, 0, 1), method = "ML")
stats 		ARMA estimation using ML
Differs from Stata: Fixed workflow; adapt the named data columns explicitly. Explicit likelihood setting avoids relying on R's default CSS-ML initialization.
Stata reference · R reference 
arima y, arima(1,0,1) // conditional least-squares alternative
 
stats::arima(y, order = c(1, 0, 1), method = "CSS")
stats 		ARMA estimation using CSS
Fixed workflow; adapt the named data columns explicitly. Conditional sum of squares is a workflow alternative, not Stata's default likelihood.
Stata reference · R reference 
arima y, arima(1,1,0)
 
stats::arima(y, order = c(1, 1, 0))
stats 		ARIMA with first difference
Differs from Stata: Fixed workflow; adapt the named data columns explicitly. Parameterization, initialization and likelihood conventions differ; compare selected model, not default output.
Stata reference · R reference 
arimasoc y, maxar(2) maxma(2)
 
{ grid <- expand.grid(p = 0:2, q = 0:2); grid$AIC <- mapply(function(p, q) stats::AIC(stats::arima(y, order = c(p, 0, q), method = "ML")), grid$p, grid$q); grid }
stats 		ARMA information-criterion grid
Fixed workflow; adapt the named data columns explicitly. Comparable-sample model-selection workflow; R and Stata likelihood constants may differ.
Stata reference · R reference 
corrgram y, lags(8)
 
list(ac = stats::acf(y, lag.max = 8, plot = FALSE), pac = stats::pacf(y, lag.max = 8, plot = FALSE))
stats 		Autocorrelation and partial-autocorrelation table
Differs from Stata: Fixed workflow; adapt the named data columns explicitly. R returns numeric correlations; request plotting separately.
Stata reference · R reference 
cumsp y
 
{ sp <- stats::spectrum(y, plot = FALSE, taper = 0); data.frame(frequency = sp$freq, cumulative = cumsum(sp$spec) / sum(sp$spec)) }
stats 		Cumulative spectral distribution
Fixed workflow; adapt the named data columns explicitly. Computes curve data; confidence bands are not included.
Stata reference · R reference 
dfgls y, maxlag(4)
 
urca::ur.ers(y, type = "DF-GLS", model = "trend", lag.max = 4)
urca 		DF-GLS unit-root test with trend
Differs from Stata: Fixed workflow; adapt the named data columns explicitly. Lag-selection and finite-sample critical values may differ.
Stata reference · R reference 
dfgls y, maxlag(4) notrend
 
urca::ur.ers(y, type = "DF-GLS", model = "constant", lag.max = 4)
urca 		DF-GLS unit-root test with constant
Differs from Stata: Fixed workflow; adapt the named data columns explicitly. Lag-selection and finite-sample critical values may differ.
Stata reference · R reference 
dfuller y, lags(0)
 
urca::ur.df(y, type = "drift", lags = 0)
urca 		ADF unit-root test with drift and 0 augmentation lags
Differs from Stata: Fixed workflow; adapt the named data columns explicitly. Match deterministic terms, sample window and lag selection; critical-value implementations can differ.
Stata reference · R reference 
dfuller y, lags(0) noconstant
 
urca::ur.df(y, type = "none", lags = 0)
urca 		ADF unit-root test with none and 0 augmentation lags
Differs from Stata: Fixed workflow; adapt the named data columns explicitly. Match deterministic terms, sample window and lag selection; critical-value implementations can differ.
Stata reference · R reference 
dfuller y, lags(0) trend
 
urca::ur.df(y, type = "trend", lags = 0)
urca 		ADF unit-root test with trend and 0 augmentation lags
Differs from Stata: Fixed workflow; adapt the named data columns explicitly. Match deterministic terms, sample window and lag selection; critical-value implementations can differ.
Stata reference · R reference 
dfuller y, lags(2)
 
urca::ur.df(y, type = "drift", lags = 2)
urca 		ADF unit-root test with drift and 2 augmentation lags
Differs from Stata: Fixed workflow; adapt the named data columns explicitly. Match deterministic terms, sample window and lag selection; critical-value implementations can differ.
Stata reference · R reference 
dfuller y, lags(2) noconstant
 
urca::ur.df(y, type = "none", lags = 2)
urca 		ADF unit-root test with none and 2 augmentation lags
Differs from Stata: Fixed workflow; adapt the named data columns explicitly. Match deterministic terms, sample window and lag selection; critical-value implementations can differ.
Stata reference · R reference 
dfuller y, lags(2) trend
 
urca::ur.df(y, type = "trend", lags = 2)
urca 		ADF unit-root test with trend and 2 augmentation lags
Differs from Stata: Fixed workflow; adapt the named data columns explicitly. Match deterministic terms, sample window and lag selection; critical-value implementations can differ.
Stata reference · R reference 
fcast compute pred, step(8)
 
{ out <- matrix(NA_real_, 8, 2); last <- tail(Y, 1); intercept <- stats::coef(fit)[1, ]; for (h in 1:8) { last <- matrix(intercept + drop(A %*% drop(last)), nrow = 1); out[h, ] <- last }; colnames(out) <- colnames(Y); out }
stats 		Iterated bivariate VAR point forecasts
Differs from Stata: Fixed workflow; adapt the named data columns explicitly. Point forecasts only; no system forecast-error covariance or interval calculation.
Stata reference · R reference 
forecast solve, begin(161)
 
forecast::forecast(fit, h = 8)
forecast 		Eight-period ARIMA forecast
Fixed workflow; adapt the named data columns explicitly. Single-equation forecasting workflow; forecast system identities are not translated.
Stata reference · R reference 
forecast solve, begin(161) periods(8) simulate(errors, statistic(stddev, prefix(sd_)))
 
forecast::forecast(fit, h = 8, level = 90)
forecast 		Forecast with ninety-percent prediction intervals
Fixed workflow; adapt the named data columns explicitly. Single-equation forecasting workflow; forecast system identities are not translated. R produces analytic 90% prediction intervals; the Stata example instead simulates forecast standard deviations, requiring a separate interval calculation.
Stata reference · R reference 
forecast solve, begin(161) static
 
stats::fitted(forecast::Arima(y, order = c(1, 0, 0)))
forecast 		One-step fitted values from ARIMA
Fixed workflow; adapt the named data columns explicitly. In-sample one-step prediction workflow; Stata's forecast model setup is separate.
Stata reference · R reference 
irf create model, step(8)
 
{ powers <- vector("list", 9); powers[[1]] <- diag(2); for (h in 2:9) powers[[h]] <- powers[[h - 1]] %*% A; powers }
base 		Reduced-form VAR(1) impulse-response matrices
Differs from Stata: Fixed workflow; adapt the named data columns explicitly. Unit reduced-form innovation responses; orthogonalized structural responses require an identification scheme.
Stata reference · R reference 
irf create model, step(8) // orthogonalized responses
 
{ impact <- t(chol(crossprod(stats::residuals(fit)) / stats::nobs(fit))); powers <- vector("list", 9); powers[[1]] <- impact; for (h in 2:9) powers[[h]] <- A %*% powers[[h - 1]]; powers }
base 		Cholesky-orthogonalized VAR responses
Differs from Stata: Fixed workflow; adapt the named data columns explicitly. Recursive ordering is y then x; covariance divisor and identifying assumptions matter.
Stata reference · R reference 
ivlpirf y, endogenous(x = z) step(8) lags(1)
 
vapply(0:8, function(h) { i <- 2:(nrow(data) - h); d <- data.frame(outcome = data$y[i + h], impulse = data$x[i], Ly = data$y[i - 1], instrument = data$z[i]); unname(stats::coef(AER::ivreg(outcome ~ impulse + Ly | instrument + Ly, data = d))[2]) }, numeric(1))
AER 		Instrumental-variable local-projection point responses
Fixed workflow; adapt the named data columns explicitly. Horizonwise 2SLS workflow; instrument validity, normalization and HAC inference require separate decisions.
Stata reference · R reference 
lpirf y, exog(x) step(8) lags(1)
 
vapply(0:8, function(h) { i <- 2:(nrow(data) - h); unname(stats::coef(stats::lm(data$y[i + h] ~ data$x[i] + data$y[i - 1] + data$x[i - 1]))[2]) }, numeric(1))
stats 		Local-projection dynamic response to an external impulse
Fixed workflow; adapt the named data columns explicitly. Simple horizonwise point-estimation workflow; HAC inference and system restrictions must be added for research use.
Stata reference · R reference 
newey y x, lag(4)
 
{ fit <- stats::lm(y ~ x, data = data); lmtest::coeftest(fit, vcov. = sandwich::NeweyWest(fit, lag = 4, prewhite = FALSE, adjust = TRUE)) }
sandwich 		Newey-West HAC coefficient inference
Differs from Stata: Fixed workflow; adapt the named data columns explicitly. Explicit bandwidth, prewhitening and finite-sample correction; align those choices.
Stata reference · R reference 
pac y, lags(12)
 
stats::pacf(y, lag.max = 12, plot = FALSE)
stats 		Partial autocorrelation function
Differs from Stata: Fixed workflow; adapt the named data columns explicitly. R returns numeric correlations; request plotting separately.
Stata reference · R reference 
pergram y
 
stats::spectrum(y, plot = FALSE, taper = 0, detrend = FALSE)
stats 		Raw periodogram
Differs from Stata: Fixed workflow; adapt the named data columns explicitly. Normalization and default tapering can differ.
Stata reference · R reference 
pperron y, lags(4)
 
urca::ur.pp(y, type = "Z-tau", model = "constant", use.lag = 4)
urca 		Phillips-Perron Z-tau test with constant
Differs from Stata: Fixed workflow; adapt the named data columns explicitly. Stata reports several statistics together; R selects one statistic. Kernel and deterministic terms must be matched.
Stata reference · R reference 
pperron y, lags(4)
 
urca::ur.pp(y, type = "Z-alpha", model = "constant", use.lag = 4)
urca 		Phillips-Perron Z-alpha test with constant
Differs from Stata: Fixed workflow; adapt the named data columns explicitly. Stata reports several statistics together; R selects one statistic. Kernel and deterministic terms must be matched.
Stata reference · R reference 
pperron y, lags(4) trend
 
urca::ur.pp(y, type = "Z-tau", model = "trend", use.lag = 4)
urca 		Phillips-Perron Z-tau test with trend
Differs from Stata: Fixed workflow; adapt the named data columns explicitly. Stata reports several statistics together; R selects one statistic. Kernel and deterministic terms must be matched.
Stata reference · R reference 
pperron y, lags(4) trend
 
urca::ur.pp(y, type = "Z-alpha", model = "trend", use.lag = 4)
urca 		Phillips-Perron Z-alpha test with trend
Differs from Stata: Fixed workflow; adapt the named data columns explicitly. Stata reports several statistics together; R selects one statistic. Kernel and deterministic terms must be matched.
Stata reference · R reference 
predict yhat, y dynamic(161)
 
stats::predict(fit, n.ahead = 8)
stats 		Dynamic ARIMA forecast and standard errors
Fixed workflow; adapt the named data columns explicitly. Single-equation forecasting workflow; forecast system identities are not translated.
Stata reference · R reference 
arima y, arima(1,0,0)
psdensity density, n(64)
 
{ phi <- unname(fit$coef["ar1"]); omega <- seq(0, pi, length.out = 64); data.frame(frequency = omega / (2 * pi), density = fit$sigma2 / (2 * pi * Mod(1 - phi * exp(-1i * omega))^2)) }
stats 		Parametric AR(1) spectral density
Differs from Stata: Fixed workflow; adapt the named data columns explicitly. Frequency and density normalization conventions must be reconciled.
Stata reference · R reference 
rolling _b, window(40) recursive: regress y x
 
t(vapply(40:nrow(data), function(i) stats::coef(stats::lm(y ~ x, data[seq_len(i), ])), numeric(2)))
stats 		Expanding-window recursive regression
Differs from Stata: Fixed workflow; adapt the named data columns explicitly. Uses a fixed first observation and increasing estimation endpoint.
Stata reference · R reference 
rolling _b, window(40): regress y x
 
zoo::rollapply(data, width = 40, FUN = function(z) stats::coef(stats::lm(y ~ x, data = as.data.frame(z))), by.column = FALSE, align = "right")
zoo 		Rolling-window regression coefficients
Differs from Stata: Fixed workflow; adapt the named data columns explicitly. Output rows correspond to complete right-aligned windows; failures need explicit handling in production.
Stata reference · R reference 
rolling sd=r(sd), window(20): summarize y
 
zoo::rollapply(y, width = 20, FUN = stats::sd, align = "right")
zoo 		Rolling sample volatility
Differs from Stata: Fixed workflow; adapt the named data columns explicitly. Sample standard deviation, not model-implied conditional volatility.
Stata reference · R reference 
sspace (level L.level L.slope, state noconstant) (slope L.slope, state noconstant) (y level, noconstant)
 
stats::StructTS(y, type = "trend")
stats 		Local-linear-trend state-space model
Differs from Stata: Fixed workflow; adapt the named data columns explicitly. Specific local-linear-trend model; initialization can differ.
Stata reference · R reference 
sspace (level L.level, state noconstant) (y level, noconstant)
 
stats::StructTS(y, type = "level")
stats 		Local-level state-space model
Differs from Stata: Fixed workflow; adapt the named data columns explicitly. Specific local-level model equivalent; sspace supports a much wider model class.
Stata reference · R reference 
tsfilter bk cycle = y, minperiod(6) maxperiod(32) smaorder(12)
 
{ K <- 12; j <- -K:K; low <- 2 * pi / 32; high <- 2 * pi / 6; w <- ifelse(j == 0, (high - low) / pi, (sin(high * j) - sin(low * j)) / (pi * j)); w <- w - mean(w); stats::filter(y, w, sides = 2) }
stats 		Baxter-King approximate band-pass cycle
Differs from Stata: Fixed workflow; adapt the named data columns explicitly. Symmetric finite-order band-pass filter loses endpoint observations.
Stata reference · R reference 
tsfilter hp cycle = y, smooth(1600) trend(trend)
 
{ n <- length(y); D <- diff(diag(n), differences = 2); trend <- drop(solve(diag(n) + 1600 * crossprod(D), as.numeric(y))); data.frame(trend = trend, cycle = as.numeric(y) - trend) }
base 		Hodrick-Prescott trend and cycle decomposition
Differs from Stata: Fixed workflow; adapt the named data columns explicitly. Complete regular series; this direct penalized least-squares solution uses the two-sided HP filter.
Stata reference · R reference 
tsreport, report
 
data.frame(previous = head(data$t, -1), next_time = tail(data$t, -1), gap = diff(data$t))[diff(data$t) != 1, ]
base 		Locate missing time periods
Fixed workflow; adapt the named data columns explicitly. Assumes integer unit-spaced time; date calendars need the intended interval definition.
Stata reference · R reference 
tssmooth exponential smooth = y, parms(.3)
 
stats::HoltWinters(y, alpha = .3, beta = FALSE, gamma = FALSE)
stats 		Simple exponential smoothing
Fixed workflow; adapt the named data columns explicitly. Initial values, endpoints and trend formulations differ; align before comparing.
Stata reference · R reference 
tssmooth hwinters smooth = y
 
stats::HoltWinters(y, gamma = FALSE)
stats 		Holt trend smoothing workflow
Fixed workflow; adapt the named data columns explicitly. Initial values, endpoints and trend formulations differ; align before comparing.
Stata reference · R reference 
tssmooth ma smooth = y, window(2 1 2)
 
stats::filter(y, rep(1 / 5, 5), sides = 2)
stats 		Centered five-term moving average
Fixed workflow; adapt the named data columns explicitly. Initial values, endpoints and trend formulations differ; align before comparing.
Stata reference · R reference 
tssmooth ma smooth = y, window(3 1 0)
 
stats::filter(y, rep(1 / 4, 4), sides = 1)
stats 		Trailing four-term moving average
Fixed workflow; adapt the named data columns explicitly. Initial values, endpoints and trend formulations differ; align before comparing.
Stata reference · R reference 
tssmooth nl smooth = y, smoother(3)
 
stats::runmed(as.numeric(y), k = 3, endrule = "median")
stats 		Running-median nonlinear smoother
Differs from Stata: Fixed workflow; adapt the named data columns explicitly. Endpoint rules should be checked separately before comparing.
Stata reference · R reference 
tssmooth shwinters smooth = y
 
stats::HoltWinters(y, seasonal = "additive")
stats 		Additive seasonal Holt-Winters smoothing
Fixed workflow; adapt the named data columns explicitly. Initial values, endpoints and trend formulations differ; align before comparing.
Stata reference · R reference 
ucm y, model(llevel)
 
stats::StructTS(y, type = "level")
stats 		Unobserved components: level
Differs from Stata: Fixed workflow; adapt the named data columns explicitly. R supports selected structural models; Stata ucm permits additional components and constraints.
Stata reference · R reference 
ucm y, model(lltrend)
 
stats::StructTS(y, type = "trend")
stats 		Unobserved components: trend
Differs from Stata: Fixed workflow; adapt the named data columns explicitly. R supports selected structural models; Stata ucm permits additional components and constraints.
Stata reference · R reference 
ucm y, model(lltrend) seasonal(4)
 
stats::StructTS(y, type = "BSM")
stats 		Unobserved components: BSM
Differs from Stata: Fixed workflow; adapt the named data columns explicitly. R supports selected structural models; Stata ucm permits additional components and constraints.
Stata reference · R reference 
var y x, lags(1)
 
stats::lm(cbind(y, x) ~ Ly + Lx, data = lagged)
stats 		Unrestricted bivariate VAR(1) by equation OLS
Differs from Stata: Fixed workflow; adapt the named data columns explicitly. Coefficient equivalent for an unrestricted common-sample VAR; system covariance reporting differs.
Stata reference · R reference 
var y x, lags(1) exog(trend)
 
stats::lm(cbind(y, x) ~ Ly + Lx + seq_len(nrow(lagged)), data = lagged)
stats 		VAR with deterministic time trend
Differs from Stata: Fixed workflow; adapt the named data columns explicitly. Trend origin affects intercepts; compare the same sample and trend indexing.
Stata reference · R reference 
var y x, lags(1) noconstant
 
stats::lm(cbind(y, x) ~ 0 + Ly + Lx, data = lagged)
stats 		Zero-intercept bivariate VAR
Differs from Stata: Fixed workflow; adapt the named data columns explicitly. Specify identical lag matrices and deterministic terms.
Stata reference · R reference 
var y x, lags(1)
vargranger
 
lmtest::grangertest(y ~ x, order = 1, data = data)
lmtest 		Test whether x Granger-predicts y
Differs from Stata: Fit var y x, lags(1) before vargranger; Stata VAR otherwise defaults to lags(1 2). grangertest uses an F reference distribution, whereas vargranger reports a Wald chi-square; match direction and sample, and do not interpret predictability as structural causation.
Stata reference · R reference 
var y x, lags(1)
vargranger
 
lmtest::grangertest(x ~ y, order = 1, data = data)
lmtest 		Test whether y Granger-predicts x
Differs from Stata: Fit var y x, lags(1) before vargranger; Stata VAR otherwise defaults to lags(1 2). grangertest uses an F reference distribution, whereas vargranger reports a Wald chi-square; match direction and sample, and do not interpret predictability as structural causation.
Stata reference · R reference 
varstable
 
eigen(A, only.values = TRUE)$values
base 		Eigenvalues of the VAR(1) companion matrix
Differs from Stata: Fixed workflow; adapt the named data columns explicitly. Stability requires all companion roots strictly inside the unit circle.
Stata reference · R reference 
vec y x, rank(1) lags(2)
 
urca::cajorls(urca::ca.jo(Y, type = "trace", ecdet = "none", K = 2, spec = "transitory"), r = 1)
urca 		Rank-one vector error-correction model
Differs from Stata: urca ecdet="const" matches Stata trend(rconstant); ecdet="none" includes an unrestricted intercept and matches trend(constant); ecdet="trend" matches trend(rtrend). ca.jo does not provide Stata trend(none) or trend(trend). Match cointegrating-vector normalization before comparing coefficients.
Stata reference · R reference 
vecrank y x, lags(2) trend(constant)
 
urca::ca.jo(Y, type = "trace", ecdet = "none", K = 2, spec = "transitory")
urca 		Test cointegration rank with an unrestricted intercept
Differs from Stata: urca ecdet="const" matches Stata trend(rconstant); ecdet="none" includes an unrestricted intercept and matches trend(constant); ecdet="trend" matches trend(rtrend). ca.jo does not provide Stata trend(none) or trend(trend). Match cointegrating-vector normalization before comparing coefficients.
Stata reference · R reference 
vecrank y x, lags(2) trend(constant) max
 
urca::ca.jo(Y, type = "eigen", ecdet = "none", K = 2, spec = "transitory")
urca 		Test cointegration rank with an unrestricted intercept
Differs from Stata: urca ecdet="const" matches Stata trend(rconstant); ecdet="none" includes an unrestricted intercept and matches trend(constant); ecdet="trend" matches trend(rtrend). ca.jo does not provide Stata trend(none) or trend(trend). Match cointegrating-vector normalization before comparing coefficients.
Stata reference · R reference 
vecrank y x, lags(2) trend(rconstant)
 
urca::ca.jo(Y, type = "trace", ecdet = "const", K = 2, spec = "transitory")
urca 		Johansen trace rank test with a restricted constant
Differs from Stata: urca ecdet="const" matches Stata trend(rconstant); ecdet="none" includes an unrestricted intercept and matches trend(constant); ecdet="trend" matches trend(rtrend). ca.jo does not provide Stata trend(none) or trend(trend). Match cointegrating-vector normalization before comparing coefficients.
Stata reference · R reference 
vecrank y x, lags(2) trend(rconstant) max
 
urca::ca.jo(Y, type = "eigen", ecdet = "const", K = 2, spec = "transitory")
urca 		Johansen maximum-eigenvalue rank test with a restricted constant
Differs from Stata: urca ecdet="const" matches Stata trend(rconstant); ecdet="none" includes an unrestricted intercept and matches trend(constant); ecdet="trend" matches trend(rtrend). ca.jo does not provide Stata trend(none) or trend(trend). Match cointegrating-vector normalization before comparing coefficients.
Stata reference · R reference 
vecrank y x, lags(2) trend(rtrend)
 
urca::ca.jo(Y, type = "trace", ecdet = "trend", K = 2, spec = "transitory")
urca 		Johansen trace rank test with a restricted trend
Differs from Stata: urca ecdet="const" matches Stata trend(rconstant); ecdet="none" includes an unrestricted intercept and matches trend(constant); ecdet="trend" matches trend(rtrend). ca.jo does not provide Stata trend(none) or trend(trend). Match cointegrating-vector normalization before comparing coefficients.
Stata reference · R reference 
vecrank y x, lags(2) trend(rtrend) max
 
urca::ca.jo(Y, type = "eigen", ecdet = "trend", K = 2, spec = "transitory")
urca 		Johansen maximum-eigenvalue rank test with a restricted trend
Differs from Stata: urca ecdet="const" matches Stata trend(rconstant); ecdet="none" includes an unrestricted intercept and matches trend(constant); ecdet="trend" matches trend(rtrend). ca.jo does not provide Stata trend(none) or trend(trend). Match cointegrating-vector normalization before comparing coefficients.
Stata reference · R reference 
predict e, residuals
wntestq e, lags(8)
 
stats::Box.test(stats::residuals(fit), lag = 8, type = "Ljung-Box", fitdf = 1)
stats 		Ljung-Box test after AR(1) estimation
Differs from Stata: Fixed workflow; adapt the named data columns explicitly. R explicitly deducts the estimated AR parameter from the reference degrees of freedom.
Stata reference · R reference 
wntestq y, lags(8)
 
stats::Box.test(y, lag = 8, type = "Ljung-Box")
stats 		Ljung-Box serial-correlation portmanteau test
Differs from Stata: Fixed workflow; adapt the named data columns explicitly. Use fitdf for tests on estimated-model residuals.
Stata reference · R reference 
xcorr y x, lags(8)
 
stats::ccf(y, x, lag.max = 8, plot = FALSE)
stats 		Cross-correlation at positive and negative lags
Differs from Stata: Fixed workflow; adapt the named data columns explicitly. Check lag sign conventions before interpreting lead-lag direction.
Stata reference · R reference

Complex survey analysis

Stata R Purpose, qualifications and sources 
estat cv
 
survey::cv(survey::svymean(~y, des))
survey 		Coefficient of variation of a survey mean
Differs from Stata: Fixed workflow; adapt the named data columns explicitly. Relative standard error is unstable when the estimated mean is near zero.
Stata reference · R reference 
estat effects
 
survey::deff(survey::svymean(~y, des, deff = TRUE))
survey 		Design effect of a survey mean
Differs from Stata: Fixed workflow; adapt the named data columns explicitly. Design effects depend on the SRS comparison, finite-population correction and weight scaling.
Stata reference · R reference 
svy bootstrap: mean y
 
{ reps <- survey::as.svrepdesign(des, type = "bootstrap"); survey::svymean(~y, reps) }
survey 		Survey mean with bootstrap replicate variance
Differs from Stata: Fixed workflow; adapt the named data columns explicitly. Point means agree with sampling weights; replication variance depends on the chosen scheme.
Stata reference · R reference 
svy brr: mean y
 
{ reps <- survey::as.svrepdesign(des, type = "BRR"); survey::svymean(~y, reps) }
survey 		Survey mean with BRR replicate variance
Differs from Stata: Fixed workflow; adapt the named data columns explicitly. Point means agree with sampling weights; replication variance depends on the chosen scheme.
Stata reference · R reference 
svy jackknife: mean y
 
{ reps <- survey::as.svrepdesign(des, type = "JKn"); survey::svymean(~y, reps) }
survey 		Survey mean with JKn replicate variance
Differs from Stata: Fixed workflow; adapt the named data columns explicitly. Point means agree with sampling weights; replication variance depends on the chosen scheme.
Stata reference · R reference 
svy, subpop(if group==1): mean y
 
survey::svymean(~y, subset(des, group == "1"))
survey 		Survey subpopulation mean
Differs from Stata: Fixed workflow; adapt the named data columns explicitly. Subset the design, preserving the sampling structure, rather than rebuilding a design after filtering.
Stata reference · R reference 
svy: glm y x, family(gamma) link(log)
 
survey::svyglm(y ~ x, des, family = stats::Gamma(link = "log"))
survey 		Survey gamma log-link model
Differs from Stata: Fixed workflow; adapt the named data columns explicitly. Quasi families retain the mean/variance structure without non-integer count warnings; align design degrees of freedom.
Stata reference · R reference 
svy: logit yes x
 
survey::svyglm(yes ~ x, des, family = stats::quasibinomial())
survey 		Survey logistic regression
Differs from Stata: Fixed workflow; adapt the named data columns explicitly. Quasi families retain the mean/variance structure without non-integer count warnings; align design degrees of freedom.
Stata reference · R reference 
svy: mean y
 
survey::svymean(~y, des)
survey 		Survey-weighted mean
Differs from Stata: Fixed workflow; adapt the named data columns explicitly. Design-based variance uses the full sampling design; population totals require correctly scaled weights.
Stata reference · R reference 
svy: mean y x
 
survey::svymean(~y + x, des)
survey 		Joint means and covariance matrix
Differs from Stata: Fixed workflow; adapt the named data columns explicitly. Design-based variance uses the full sampling design; population totals require correctly scaled weights.
Stata reference · R reference 
svy: mean y, over(group)
 
survey::svyby(~y, ~group, des, survey::svymean)
survey 		Domain-specific means
Differs from Stata: Fixed workflow; adapt the named data columns explicitly. Design-based variance uses the full sampling design; population totals require correctly scaled weights.
Stata reference · R reference 
svy: poisson y x
 
survey::svyglm(y ~ x, des, family = stats::quasipoisson())
survey 		Survey Poisson mean model
Differs from Stata: Fixed workflow; adapt the named data columns explicitly. Quasi families retain the mean/variance structure without non-integer count warnings; align design degrees of freedom.
Stata reference · R reference 
svy: probit yes x
 
survey::svyglm(yes ~ x, des, family = stats::quasibinomial(link = "probit"))
survey 		Survey probit regression
Differs from Stata: Fixed workflow; adapt the named data columns explicitly. Quasi families retain the mean/variance structure without non-integer count warnings; align design degrees of freedom.
Stata reference · R reference 
svy: proportion yes
 
survey::svymean(~factor(yes), des)
survey 		Survey-weighted category proportions
Differs from Stata: Fixed workflow; adapt the named data columns explicitly. Design-based variance uses the full sampling design; population totals require correctly scaled weights.
Stata reference · R reference 
svy: proportion yes
 
survey::svyciprop(~yes, des, method = "logit", level = .95)
survey 		Survey proportion interval using logit
Fixed workflow; adapt the named data columns explicitly. Different interval constructions are intentional alternatives; match Stata's selected transformation.
Stata reference · R reference 
svy: proportion yes
 
survey::svyciprop(~yes, des, method = "likelihood", level = .95)
survey 		Survey proportion interval using likelihood
Fixed workflow; adapt the named data columns explicitly. Different interval constructions are intentional alternatives; match Stata's selected transformation.
Stata reference · R reference 
svy: proportion yes
 
survey::svyciprop(~yes, des, method = "asin", level = .95)
survey 		Survey proportion interval using asin
Fixed workflow; adapt the named data columns explicitly. Different interval constructions are intentional alternatives; match Stata's selected transformation.
Stata reference · R reference 
svy: proportion yes
 
survey::svyciprop(~yes, des, method = "beta", level = .95)
survey 		Survey proportion interval using beta
Fixed workflow; adapt the named data columns explicitly. Different interval constructions are intentional alternatives; match Stata's selected transformation.
Stata reference · R reference 
svy: proportion yes
 
survey::svyciprop(~yes, des, method = "mean", level = .95)
survey 		Survey proportion interval using mean
Fixed workflow; adapt the named data columns explicitly. Different interval constructions are intentional alternatives; match Stata's selected transformation.
Stata reference · R reference 
svy: proportion yes
 
survey::svyciprop(~yes, des, method = "xlogit", level = .95)
survey 		Survey proportion interval using xlogit
Fixed workflow; adapt the named data columns explicitly. Different interval constructions are intentional alternatives; match Stata's selected transformation.
Stata reference · R reference 
svy: proportion yes, over(group)
 
survey::svyby(~factor(yes), ~group, des, survey::svymean)
survey 		Domain-specific proportions
Differs from Stata: Fixed workflow; adapt the named data columns explicitly. Design-based variance uses the full sampling design; population totals require correctly scaled weights.
Stata reference · R reference 
svy: ratio y/x
 
survey::svyratio(~y, ~x, des)
survey 		Survey ratio of totals
Differs from Stata: Fixed workflow; adapt the named data columns explicitly. Design-based variance uses the full sampling design; population totals require correctly scaled weights.
Stata reference · R reference 
svy: regress y x
 
survey::svyglm(y ~ x, des, family = stats::gaussian())
survey 		Survey linear regression
Differs from Stata: Fixed workflow; adapt the named data columns explicitly. Quasi families retain the mean/variance structure without non-integer count warnings; align design degrees of freedom.
Stata reference · R reference 
svy: tabulate group yes
 
prop.table(survey::svytable(~group + yes, des))
survey 		Weighted contingency-table counts
Differs from Stata: Returns weighted cell proportions, the default display of svy: tabulate twoway. This table alone does not provide Stata design-based standard errors, confidence intervals or adjusted association tests.
Stata reference · R reference 
svy: tabulate group yes, column
 
prop.table(survey::svytable(~group + yes, des), margin = 2)
survey 		Survey column proportions
Differs from Stata: Fixed workflow; adapt the named data columns explicitly. Design-based variance uses the full sampling design; population totals require correctly scaled weights.
Stata reference · R reference 
svy: tabulate group yes, pearson
 
survey::svychisq(~group + yes, des, statistic = "F")
survey 		Design-based categorical association: F
Fixed workflow; adapt the named data columns explicitly. Rao-Scott and Wald choices have distinct reference distributions; choose the intended test explicitly.
Stata reference · R reference 
svy: tabulate group yes, pearson
 
survey::svychisq(~group + yes, des, statistic = "Chisq")
survey 		Design-based categorical association: Chisq
Fixed workflow; adapt the named data columns explicitly. Rao-Scott and Wald choices have distinct reference distributions; choose the intended test explicitly.
Stata reference · R reference 
svy: tabulate group yes, pearson
 
survey::svychisq(~group + yes, des, statistic = "Wald")
survey 		Design-based categorical association: Wald
Fixed workflow; adapt the named data columns explicitly. Rao-Scott and Wald choices have distinct reference distributions; choose the intended test explicitly.
Stata reference · R reference 
svy: tabulate group yes, pearson
 
survey::svychisq(~group + yes, des, statistic = "adjWald")
survey 		Design-based categorical association: adjWald
Fixed workflow; adapt the named data columns explicitly. Rao-Scott and Wald choices have distinct reference distributions; choose the intended test explicitly.
Stata reference · R reference 
svy: tabulate group yes, row
 
prop.table(survey::svytable(~group + yes, des), margin = 1)
survey 		Survey row proportions
Differs from Stata: Fixed workflow; adapt the named data columns explicitly. Design-based variance uses the full sampling design; population totals require correctly scaled weights.
Stata reference · R reference 
svy: total y
 
survey::svytotal(~y, des)
survey 		Survey-weighted population total
Differs from Stata: Fixed workflow; adapt the named data columns explicitly. Design-based variance uses the full sampling design; population totals require correctly scaled weights.
Stata reference · R reference 
svy: total y x
 
survey::svytotal(~y + x, des)
survey 		Joint population totals
Differs from Stata: Fixed workflow; adapt the named data columns explicitly. Design-based variance uses the full sampling design; population totals require correctly scaled weights.
Stata reference · R reference 
svy: total y, over(group)
 
survey::svyby(~y, ~group, des, survey::svytotal)
survey 		Domain-specific population totals
Differs from Stata: Fixed workflow; adapt the named data columns explicitly. Design-based variance uses the full sampling design; population totals require correctly scaled weights.
Stata reference · R reference 
svydescribe
 
list(degrees_of_freedom = survey::degf(des), strata = table(des$strata[, 1]), clusters = length(unique(des$cluster[, 1])))
survey 		Survey design degrees of freedom and cluster counts
Differs from Stata: Fixed workflow; adapt the named data columns explicitly. Counts observed records by stratum; PSU count is distinct from observation count.
Stata reference · R reference 
svyset [pw=w]
 
survey::svydesign(ids = ~1, weights = ~w, data = data)
survey 		Independent unequal-probability sample
Differs from Stata: Fixed workflow; adapt the named data columns explicitly. Survey objects carry design metadata; do not substitute ordinary regression weights.
Stata reference · R reference 
svyset psu [pw=w], strata(stratum)
 
survey::svydesign(ids = ~psu, strata = ~stratum, weights = ~w, data = data)
survey 		Stratified cluster survey design
Differs from Stata: Fixed workflow; adapt the named data columns explicitly. Survey objects carry design metadata; do not substitute ordinary regression weights.
Stata reference · R reference 
svyset psu [pw=w], strata(stratum) fpc(N)
 
survey::svydesign(ids = ~psu, strata = ~stratum, weights = ~w, fpc = ~N, data = data)
survey 		Finite-population correction in a clustered design
Differs from Stata: Fixed workflow; adapt the named data columns explicitly. Survey objects carry design metadata; do not substitute ordinary regression weights.
Stata reference · R reference 
svyset psu [pw=w], strata(stratum) poststrata(group) postweight(pop)
 
survey::postStratify(des, ~group, population = population)
survey 		Poststratify survey weights to population category totals
Differs from Stata: Fixed workflow; adapt the named data columns explicitly. Population categories must cover the sample; weighting adjustments are part of the variance design.
Stata reference · R reference 
svyset psu [pw=w], strata(stratum) rake(ibn.group, totals(20 30, copy) noconstant)
 
survey::rake(des, sample.margins = list(~group), population.margins = list(population))
survey 		Rake sample weights to known category margins
Fixed workflow; adapt the named data columns explicitly. Survey calibration workflow; Stata's exact calibration syntax depends on svyset options/version.
Stata reference · R reference 
svyset psu [pw=w], strata(stratum) vce(bootstrap)
 
survey::as.svrepdesign(des, type = "bootstrap", replicates = 50)
survey 		Survey bootstrap replicate-weight design
Fixed workflow; adapt the named data columns explicitly. Illustrative replication count; specific survey bootstrap schemes and scales can differ.
Stata reference · R reference 
svyset psu [pw=w], strata(stratum) vce(brr)
 
survey::as.svrepdesign(des, type = "BRR")
survey 		Balanced repeated replication for paired-PSU strata
Differs from Stata: Fixed workflow; adapt the named data columns explicitly. BRR requires exactly two PSUs per stratum unless a deliberate design adjustment is made.
Stata reference · R reference 
svyset psu [pw=w], strata(stratum) vce(brr) fay(.3)
 
survey::as.svrepdesign(des, type = "Fay", fay.rho = .3)
survey 		Fay balanced repeated replication
Differs from Stata: Fixed workflow; adapt the named data columns explicitly. Match Fay's perturbation parameter and replicate scaling conventions.
Stata reference · R reference 
svyset psu [pw=w], vce(jackknife)
 
survey::as.svrepdesign(des, type = "JKn")
survey 		Stratified delete-one jackknife replicates
Differs from Stata: Fixed workflow; adapt the named data columns explicitly. Survey objects carry design metadata; do not substitute ordinary regression weights.
Stata reference · R reference

Meta-analysis

Stata R Purpose, qualifications and sources 
meta bias, begg
 
metafor::ranktest(metafor::rma(yi, vi, data = data, method = "FE"))
metafor 		Rank-correlation small-study-effects test
Differs from Stata: Fixed workflow; adapt the named data columns explicitly. Rank-correlation and small-sample p-value conventions may differ.
Stata reference · R reference 
meta bias, egger
 
metafor::regtest(metafor::rma(yi, vi, data = data, method = "REML"), model = "rma", predictor = "sei")
metafor 		Egger regression asymmetry test
Differs from Stata: Assumes the Stata meta declaration is its default random(reml). model="rma" regresses the effects on their standard errors with additive REML heterogeneity; model="lm" would instead use multiplicative dispersion. Small-study effects alone do not establish publication bias.
Stata reference · R reference 
meta esize a b c d, esize(lnoratio)
 
metafor::escalc(measure = "OR", ai = a, bi = b, ci = c, di = d, data = data)
metafor 		Log odds ratios and variances
Differs from Stata: Fixed workflow; adapt the named data columns explicitly. Use treatment/control cell ordering consistently and select continuity corrections deliberately.
Stata reference · R reference 
meta esize a b c d, esize(lnrratio)
 
metafor::escalc(measure = "RR", ai = a, bi = b, ci = c, di = d, data = data)
metafor 		Log risk ratios and variances
Differs from Stata: Fixed workflow; adapt the named data columns explicitly. Use treatment/control cell ordering consistently and select continuity corrections deliberately.
Stata reference · R reference 
meta esize a b c d, esize(rdiff)
 
metafor::escalc(measure = "RD", ai = a, bi = b, ci = c, di = d, data = data)
metafor 		Risk differences and variances
Differs from Stata: Fixed workflow; adapt the named data columns explicitly. Use treatment/control cell ordering consistently and select continuity corrections deliberately.
Stata reference · R reference 
meta esize cases n, esize(ftukeyprop)
 
metafor::escalc(measure = "PFT", xi = cases, ni = n, data = data)
metafor 		Freeman-Tukey transformed proportions
Differs from Stata: Fixed workflow; adapt the named data columns explicitly. Transformation definitions and back-transformation choices, especially double-arcsine, require care.
Stata reference · R reference 
meta esize cases n, esize(logitprop)
 
metafor::escalc(measure = "PLO", xi = cases, ni = n, data = data)
metafor 		Logit-transformed study proportions
Differs from Stata: Fixed workflow; adapt the named data columns explicitly. Transformation definitions and back-transformation choices, especially double-arcsine, require care.
Stata reference · R reference 
meta esize cases n, esize(proportion)
 
metafor::escalc(measure = "PR", xi = cases, ni = n, data = data)
metafor 		Raw study proportions and variances
Differs from Stata: Fixed workflow; adapt the named data columns explicitly. Transformation definitions and back-transformation choices, especially double-arcsine, require care.
Stata reference · R reference 
meta esize n1 m1 sd1 n0 m0 sd0, esize(cohend)
 
metafor::escalc(measure = "SMD", m1i = m1, sd1i = sd1, n1i = n1, m2i = m0, sd2i = sd0, n2i = n0, data = data, correct = FALSE)
metafor 		Uncorrected Cohen standardized mean differences
Differs from Stata: Fixed workflow; adapt the named data columns explicitly. Sampling-variance estimators for standardized effects must be matched.
Stata reference · R reference 
meta esize n1 m1 sd1 n0 m0 sd0, esize(hedgesg)
 
metafor::escalc(measure = "SMD", m1i = m1, sd1i = sd1, n1i = n1, m2i = m0, sd2i = sd0, n2i = n0, data = data)
metafor 		Bias-corrected standardized mean differences
Differs from Stata: metafor SMD uses the exact gamma-function Hedges correction. Stata hedgesg uses its approximation by default; specify esize(hedgesg, exact) to align that correction, and also match the sampling-variance formula.
Stata reference · R reference 
meta esize n1 m1 sd1 n0 m0 sd0, esize(mdiff)
 
metafor::escalc(measure = "MD", m1i = m1, sd1i = sd1, n1i = n1, m2i = m0, sd2i = sd0, n2i = n0, data = data, vtype = "HO")
metafor 		Unstandardized mean differences
Differs from Stata: vtype="HO" uses the pooled within-group variance, matching Stata esize(mdiff) default equal-variance sampling variance. metafor default vtype="LS" instead uses separate group variances; use it with Stata unequal.
Stata reference · R reference 
meta esize r n, correlation
 
metafor::escalc(measure = "COR", ri = r, ni = n, data = data)
metafor 		Raw correlations and sampling variances
Differs from Stata: Fixed workflow; adapt the named data columns explicitly. Correlations must be within (-1,1); sampling variance depends on transformation.
Stata reference · R reference 
meta esize r n, fisherz
 
metafor::escalc(measure = "ZCOR", ri = r, ni = n, data = data)
metafor 		Fisher-z transformed correlations
Differs from Stata: Fixed workflow; adapt the named data columns explicitly. Correlations must be within (-1,1); sampling variance depends on transformation.
Stata reference · R reference 
meta regress c.year
 
metafor::rma(yi = yi, vi = vi, mods = ~year, data = data, method = "REML")
metafor 		Meta-regression on centered publication year
Differs from Stata: Uses year on its original scale, so the intercept refers to year zero as in the Stata command. Assumes the Stata declared model is random(REML); match its model declaration and standard-error options.
Stata reference · R reference 
meta regress i.x, random(reml)
 
metafor::rma(yi, vi, mods = ~factor(x), data = data, method = "REML")
metafor 		Categorical meta-regression with Knapp-Hartung inference
Differs from Stata: Uses ordinary normal-reference REML meta-regression, as shown by the Stata command without se(khartung). Knapp-Hartung would require an explicit option in both systems. Moderator associations are study-level, not causal subgroup effects.
Stata reference · R reference 
meta regress x
 
metafor::rma(yi = yi, vi = vi, mods = ~x, data = data, method = "REML")
metafor 		Random-effects meta-regression with moderator
Differs from Stata: Fixed workflow; adapt the named data columns explicitly. Meta-regression relates study-level summaries, not individual-level effects.
Stata reference · R reference 
meta set yi sei
 
metafor::escalc(measure = "GEN", yi = yi, sei = sei, data = data)
metafor 		Declare generic effect sizes and sampling variances
Differs from Stata: Both examples take standard errors as sei; metafor stores their squares as vi. escalc creates effect-size data but does not persist a Stata-style meta declaration, labels or a default analysis model.
Stata reference · R reference 
meta summarize, cumulative(year)
 
metafor::cumul(metafor::rma(yi, vi, data = data, method = "REML"), order = data$year)
metafor 		Cumulative fixed-effect meta-analysis by year
Differs from Stata: Assumes meta set used its default random(reml) declaration; metafor explicitly selects REML. A prior Stata common/fixed/random declaration changes the target. Match study order, missingness and heterogeneity estimation before comparing summaries.
Stata reference · R reference 
meta summarize, eform
 
exp(c(estimate = as.numeric(fit$b), lower = fit$ci.lb, upper = fit$ci.ub))
metafor 		Exponentiate a pooled log-effect and interval
Differs from Stata: Fixed workflow; adapt the named data columns explicitly. Only appropriate when effects are on a log-ratio scale.
Stata reference · R reference 
meta summarize, fixed
 
metafor::rma(yi = yi, vi = vi, data = data, method = "FE")
metafor 		Inverse-variance meta-analysis using FE
Differs from Stata: Fixed workflow; adapt the named data columns explicitly. Between-study variance estimators and confidence procedures must be matched; some R estimator names differ from Stata's names.
Stata reference · R reference 
meta summarize, leaveoneout
 
metafor::leave1out(metafor::rma(yi, vi, data = data, method = "REML"))
metafor 		Leave-one-study-out pooled estimates
Differs from Stata: Assumes meta set used its default random(reml) declaration; metafor explicitly selects REML. A prior Stata common/fixed/random declaration changes the target. Match study order, missingness and heterogeneity estimation before comparing summaries.
Stata reference · R reference 
meta summarize, random(dlaird)
 
metafor::rma(yi = yi, vi = vi, data = data, method = "DL")
metafor 		Inverse-variance meta-analysis using DL
Differs from Stata: Fixed workflow; adapt the named data columns explicitly. Between-study variance estimators and confidence procedures must be matched; some R estimator names differ from Stata's names.
Stata reference · R reference 
meta summarize, random(ebayes)
 
metafor::rma(yi = yi, vi = vi, data = data, method = "EB")
metafor 		Inverse-variance meta-analysis using EB
Differs from Stata: Fixed workflow; adapt the named data columns explicitly. Between-study variance estimators and confidence procedures must be matched; some R estimator names differ from Stata's names.
Stata reference · R reference 
meta summarize, random(hedges)
 
metafor::rma(yi = yi, vi = vi, data = data, method = "HE")
metafor 		Inverse-variance meta-analysis using HE
Differs from Stata: Fixed workflow; adapt the named data columns explicitly. Between-study variance estimators and confidence procedures must be matched; some R estimator names differ from Stata's names.
Stata reference · R reference 
meta summarize, random(mle)
 
metafor::rma(yi = yi, vi = vi, data = data, method = "ML")
metafor 		Inverse-variance meta-analysis using ML
Differs from Stata: Fixed workflow; adapt the named data columns explicitly. Between-study variance estimators and confidence procedures must be matched; some R estimator names differ from Stata's names.
Stata reference · R reference 
meta summarize, random(reml)
 
metafor::rma(yi = yi, vi = vi, data = data, method = "REML")
metafor 		Inverse-variance meta-analysis using REML
Differs from Stata: Fixed workflow; adapt the named data columns explicitly. Between-study variance estimators and confidence procedures must be matched; some R estimator names differ from Stata's names.
Stata reference · R reference 
meta summarize, random(reml)
 
metafor::rma(yi = yi, vi = vi, data = data, method = "REML", test = "z")
metafor 		Random-effects pooled inference using z
Fixed workflow; adapt the named data columns explicitly. Only knha corresponds to a Knapp-Hartung correction; z/t are explicit inference alternatives.
Stata reference · R reference 
meta summarize, random(reml) predinterval
 
stats::predict(metafor::rma(yi, vi, data = data, method = "REML"))
metafor 		Prediction interval for a new study's true effect
Differs from Stata: The true-effect prediction interval differs from the pooled-mean confidence interval. Stata uses t(K-2); metafor defaults to a normal quantile. For Stata endpoints compute mu +/- qt(.975,K-2)*sqrt(tau2+se^2); changing metafor test="t" uses K-p degrees of freedom and is not the same correction.
Stata reference · R reference 
meta summarize, random(reml) se(khartung)
 
metafor::rma(yi = yi, vi = vi, data = data, method = "REML", test = "knha")
metafor 		Random-effects pooled inference using knha
Fixed workflow; adapt the named data columns explicitly. Only knha corresponds to a Knapp-Hartung correction; z/t are explicit inference alternatives.
Stata reference · R reference 
meta summarize, random(reml) tdistribution
 
metafor::rma(yi = yi, vi = vi, data = data, method = "REML", test = "t")
metafor 		Random-effects pooled inference using t
Fixed workflow; adapt the named data columns explicitly. Only knha corresponds to a Knapp-Hartung correction; z/t are explicit inference alternatives.
Stata reference · R reference 
meta summarize, random(sjonkman)
 
metafor::rma(yi = yi, vi = vi, data = data, method = "SJ")
metafor 		Inverse-variance meta-analysis using SJ
Differs from Stata: Fixed workflow; adapt the named data columns explicitly. Between-study variance estimators and confidence procedures must be matched; some R estimator names differ from Stata's names.
Stata reference · R reference 
meta summarize, subgroup(x)
 
lapply(split(data, data$x), function(d) metafor::rma(yi, vi, data = d, method = "REML"))
metafor 		Separate random-effects estimates by study subgroup
Differs from Stata: Fixed workflow; adapt the named data columns explicitly. Fits separate heterogeneity parameters; not the same as a shared-tau subgroup model.
Stata reference · R reference 
meta trimfill
 
metafor::trimfill(metafor::rma(yi, vi, data = data, method = "REML"))
metafor 		Trim-and-fill sensitivity analysis
Differs from Stata: Assumes meta set used its default random(reml) declaration; metafor explicitly selects REML. A prior Stata common/fixed/random declaration changes the target. Match study order, missingness and heterogeneity estimation before comparing summaries.
Stata reference · R reference

Multiple imputation

Stata R Purpose, qualifications and sources 
mi convert flong
 
mice::complete(imp, action = "long", include = TRUE)
mice 		Stack imputations including original data
Fixed workflow; adapt the named data columns explicitly. R mids metadata and storage structures differ from Stata MI data styles.
Stata reference · R reference 
mi describe
 
list(imputations = imp$m, missing_by_variable = imp$nmis, methods = imp$method)
mice 		Describe multiple-imputation metadata
Fixed workflow; adapt the named data columns explicitly. R mids metadata and storage structures differ from Stata MI data styles.
Stata reference · R reference 
mi estimate, level(90): regress y x z
 
summary(mice::pool(with(imp, stats::lm(y ~ x + z))), conf.int = TRUE, conf.level = .90)
mice 		Pooled estimates with ninety-percent intervals
Differs from Stata: Fixed workflow; adapt the named data columns explicitly. Uses mice's pooled small-sample degrees-of-freedom rules.
Stata reference · R reference 
mi estimate: logit binary x z
 
mice::pool(with(imp, stats::glm(binary ~ x + z, family = stats::binomial())))
mice 		Pool logistic-regression coefficients
Differs from Stata: Fixed workflow; adapt the named data columns explicitly. Pool on the log-odds scale before exponentiating.
Stata reference · R reference 
mi estimate: mean x
 
{ completed <- mice::complete(imp, "all"); Q <- vapply(completed, function(d) mean(d$x), numeric(1)); U <- vapply(completed, function(d) stats::var(d$x) / nrow(d), numeric(1)); mice::pool.scalar(Q, U, n = nrow(data), k = 1) }
mice 		Rubin pooling of an imputed continuous mean
Differs from Stata: Fixed workflow; adapt the named data columns explicitly. Within-imputation mean variances and complete-data sample sizes are explicit.
Stata reference · R reference 
mi estimate: poisson count x z
 
mice::pool(with(imp, stats::glm(count ~ x + z, family = stats::poisson())))
mice 		Pool Poisson-regression coefficients
Differs from Stata: Fixed workflow; adapt the named data columns explicitly. Pool log-rate coefficients; assess overdispersion separately.
Stata reference · R reference 
mi estimate: regress y x z
 
mice::pool(with(imp, stats::lm(y ~ x + z)))
mice 		Rubin pooling of linear-regression coefficients
Differs from Stata: Fixed workflow; adapt the named data columns explicitly. Pools within- and between-imputation variance; complete-data degrees of freedom must be appropriate.
Stata reference · R reference 
mi estimate: regress y x z
 
{ fits <- lapply(mice::complete(imp, "all"), function(d) stats::lm(y ~ x + z, data = d)); mitools::MIcombine(lapply(fits, stats::coef), lapply(fits, stats::vcov)) }
mitools 		Pool coefficient vectors and covariance matrices directly
Differs from Stata: Fixed workflow; adapt the named data columns explicitly. Alternative low-level Rubin pooling; degrees-of-freedom assumptions should be specified for small samples.
Stata reference · R reference 
mi extract 1
 
mice::complete(imp, 1)
mice 		Extract first completed dataset
Fixed workflow; adapt the named data columns explicitly. R mids metadata and storage structures differ from Stata MI data styles.
Stata reference · R reference 
mi impute chained (logit) category (pmm, knn(5)) x z = y, add(3)
 
mice::mice(data, m = 3, maxit = 3, method = methods, seed = 131, donors = 5L, printFlag = FALSE)
mice 		Binary logistic chained-equation imputation
Differs from Stata: Fixed workflow; adapt the named data columns explicitly. Set categorical types before imputation and match the model family; random draws are not cross-software identical.
Stata reference · R reference 
mi impute chained (mlogit) category (pmm, knn(5)) x z = y, add(3)
 
mice::mice(data, m = 3, maxit = 3, method = methods, seed = 131, donors = 5L, printFlag = FALSE)
mice 		Unordered-category chained-equation imputation
Differs from Stata: Fixed workflow; adapt the named data columns explicitly. Set categorical types before imputation and match the model family; random draws are not cross-software identical.
Stata reference · R reference 
mi impute chained (ologit) category (pmm, knn(5)) x z = y, add(3)
 
mice::mice(data, m = 3, maxit = 3, method = methods, seed = 131, donors = 5L, printFlag = FALSE)
mice 		Ordered-category chained-equation imputation
Differs from Stata: Fixed workflow; adapt the named data columns explicitly. Set categorical types before imputation and match the model family; random draws are not cross-software identical.
Stata reference · R reference 
mi impute chained (pmm, knn(5)) x z = y, add(3) // restrict predictors
 
{ predictors <- mice::make.predictorMatrix(data); predictors["x", "z"] <- 0; mice::mice(data, m = 3, maxit = 3, predictorMatrix = predictors, seed = 132, donors = 5L, printFlag = FALSE) }
mice 		Customize the imputation predictor matrix
Fixed workflow; adapt the named data columns explicitly. A workflow for per-variable models; the R predictor matrix is not inferred automatically from Stata syntax.
Stata reference · R reference 
mi impute chained (pmm, knn(5)) x z = y, add(3) rseed(108)
 
mice::mice(data, m = 3, maxit = 3, method = "pmm", seed = 108, donors = 5L, printFlag = FALSE)
mice 		Chained-equation imputation using pmm
Differs from Stata: Fixed workflow; adapt the named data columns explicitly. Imputation algorithms and random draws differ; specify predictor matrix and model types deliberately.
Stata reference · R reference 
mi impute chained (regress) x z = y, add(3) rseed(108)
 
mice::mice(data, m = 3, maxit = 3, method = "norm", seed = 108, printFlag = FALSE)
mice 		Chained-equation imputation using norm
Differs from Stata: Fixed workflow; adapt the named data columns explicitly. Imputation algorithms and random draws differ; specify predictor matrix and model types deliberately.
Stata reference · R reference 
mi passive: generate square = x^2
 
lapply(mice::complete(imp, action = "all"), function(d) transform(d, square = x^2))
mice 		Derive a nonlinear variable consistently in each completed dataset
Fixed workflow; adapt the named data columns explicitly. Post-imputation derivation is appropriate when the transform is not needed as an imputation predictor; otherwise use passive imputation within the chain.
Stata reference · R reference 
mi test x
 
{ full <- with(imp, stats::lm(y ~ x + z)); reduced <- with(imp, stats::lm(y ~ z)); mice::D1(full, reduced) }
mice 		Combine repeated complete-data Wald tests
D1 pools coefficient estimates and covariance matrices for a multivariate MI Wald test, as Stata mi test does. Use the same coefficient ordering and complete-data degrees of freedom; D2 pools test statistics and is a different rule.
Stata reference · R reference 
mi test x z
 
{ full <- with(imp, stats::lm(y ~ x + z)); reduced <- with(imp, stats::lm(y ~ 1)); mice::D1(full, reduced) }
mice 		Pooled multivariate Wald comparison
Differs from Stata: Fixed workflow; adapt the named data columns explicitly. D1 combines nested-model Wald information; coefficients must have consistent ordering.
Stata reference · R reference 
mi varying
 
vapply(imp$imp, function(z) if (nrow(z)) any(apply(z, 1, function(v) length(unique(v)) > 1)) else FALSE, logical(1))
mice 		Identify variables varying across imputations
Fixed workflow; adapt the named data columns explicitly. R mids metadata and storage structures differ from Stata MI data styles.
Stata reference · R reference 
mi xeq: summarize x
 
lapply(mice::complete(imp, action = "all"), function(d) summary(d$x))
mice 		Inspect imputed-variable distributions in each dataset
Fixed workflow; adapt the named data columns explicitly. R mids metadata and storage structures differ from Stata MI data styles.
Stata reference · R reference

Multilevel models

Stata R Purpose, qualifications and sources 
estat icc
 
{ v <- lme4::VarCorr(fit); between <- as.numeric(v$id[1, 1]); between / (between + stats::sigma(fit)^2) }
lme4 		Random-intercept intraclass correlation
Differs from Stata: Fixed workflow; adapt the named data columns explicitly. ICC is constant only for a random-intercept Gaussian model; random slopes make it covariate-dependent.
Stata reference · R reference 
estat recovariance
 
as.matrix(lme4::VarCorr(fit)$id)
lme4 		Random-effects covariance matrix
Differs from Stata: Fixed workflow; adapt the named data columns explicitly. Estimated covariance conditional on the chosen random-effects structure.
Stata reference · R reference 
estat sd
 
as.data.frame(lme4::VarCorr(fit))
lme4 		Variance components as standard deviations and correlations
Differs from Stata: Fixed workflow; adapt the named data columns explicitly. Includes residual standard deviation and random-effect correlation.
Stata reference · R reference 
mecloglog binary x || id:, intpoints(1)
 
lme4::glmer(binary ~ x + (1 | id), data = data, family = stats::binomial(link = "cloglog"), nAGQ = 1, control = lme4::glmerControl(optimizer = "bobyqa"))
lme4 		Complementary-log-log random-intercept model with Laplace approximation
Differs from Stata: Fixed workflow; adapt the named data columns explicitly. Stata's integration settings must be matched; one Laplace point differs from adaptive quadrature.
Stata reference · R reference 
mecloglog binary x || id:, intpoints(9)
 
lme4::glmer(binary ~ x + (1 | id), data = data, family = stats::binomial(link = "cloglog"), nAGQ = 9, control = lme4::glmerControl(optimizer = "bobyqa"))
lme4 		Complementary-log-log random-intercept model with adaptive quadrature
Differs from Stata: Fixed workflow; adapt the named data columns explicitly. Stata's integration settings must be matched; one Laplace point differs from adaptive quadrature.
Stata reference · R reference 
meglm binary x || id:, family(binomial) link(cloglog)
 
lme4::glmer(binary ~ x + (1 | id), data = data, family = stats::binomial(link = "cloglog"), nAGQ = 7)
lme4 		Generalized mixed model with cloglog link
Differs from Stata: Fixed workflow; adapt the named data columns explicitly. Explicit family/link choice; defaults and integration accuracy differ.
Stata reference · R reference 
meglm binary x || id:, family(binomial) link(logit)
 
lme4::glmer(binary ~ x + (1 | id), data = data, family = stats::binomial(link = "logit"), nAGQ = 7)
lme4 		Generalized mixed model with logit link
Differs from Stata: Fixed workflow; adapt the named data columns explicitly. Explicit family/link choice; defaults and integration accuracy differ.
Stata reference · R reference 
meglm binary x || id:, family(binomial) link(probit)
 
lme4::glmer(binary ~ x + (1 | id), data = data, family = stats::binomial(link = "probit"), nAGQ = 7)
lme4 		Generalized mixed model with probit link
Differs from Stata: Fixed workflow; adapt the named data columns explicitly. Explicit family/link choice; defaults and integration accuracy differ.
Stata reference · R reference 
melogit binary x || id:, intpoints(1)
 
lme4::glmer(binary ~ x + (1 | id), data = data, family = stats::binomial(link = "logit"), nAGQ = 1, control = lme4::glmerControl(optimizer = "bobyqa"))
lme4 		Logistic random-intercept model with Laplace approximation
Differs from Stata: Fixed workflow; adapt the named data columns explicitly. Stata's integration settings must be matched; one Laplace point differs from adaptive quadrature.
Stata reference · R reference 
melogit binary x || id:, intpoints(9)
 
lme4::glmer(binary ~ x + (1 | id), data = data, family = stats::binomial(link = "logit"), nAGQ = 9, control = lme4::glmerControl(optimizer = "bobyqa"))
lme4 		Logistic random-intercept model with adaptive quadrature
Differs from Stata: Fixed workflow; adapt the named data columns explicitly. Stata's integration settings must be matched; one Laplace point differs from adaptive quadrature.
Stata reference · R reference 
menbreg nb x || id:
 
glmmTMB::glmmTMB(nb ~ x + (1 | id), data = data, family = glmmTMB::nbinom2())
glmmTMB 		Negative-binomial mixed model with quadratic mean variance
Differs from Stata: glmmTMB uses Laplace integration; Stata menbreg defaults to intpoints(7) adaptive quadrature and needs intpoints(1) for a comparable integration approximation. Match NB2 dispersion versus its inverse before comparing ancillary parameters; identical integration does not itself guarantee estimator parity.
Stata reference · R reference 
menl y = {A}*exp({B}*x), define(A: {a0}+{U[id]})
 
nlme::nlme(y ~ A * exp(B * x), data = data, fixed = A + B ~ 1, random = A ~ 1 | id, start = c(A = 2, B = .25))
nlme 		Nonlinear exponential growth with random asymptote coefficient
Differs from Stata: Fixed workflow; adapt the named data columns explicitly. Nonlinear mixed-model workflow; Stata menl syntax/parameter definitions must match the desired model.
Stata reference · R reference 
meologit ordinal x || id:
 
ordinal::clmm(ordinal ~ x + (1 | id), data = data, link = "logit", nAGQ = 7)
ordinal 		Random-intercept cumulative-logit model
Differs from Stata: Fixed workflow; adapt the named data columns explicitly. Threshold signs and quadrature defaults must be aligned.
Stata reference · R reference 
meoprobit ordinal x || id:
 
ordinal::clmm(ordinal ~ x + (1 | id), data = data, link = "probit", nAGQ = 7)
ordinal 		Random-intercept cumulative-probit model
Differs from Stata: Fixed workflow; adapt the named data columns explicitly. Latent residual variance is fixed by the probit link.
Stata reference · R reference 
mepoisson count x || id:, intpoints(1)
 
lme4::glmer(count ~ x + (1 | id), data = data, family = stats::poisson(link = "log"), nAGQ = 1, control = lme4::glmerControl(optimizer = "bobyqa"))
lme4 		Poisson random-intercept model with Laplace approximation
Differs from Stata: Fixed workflow; adapt the named data columns explicitly. Stata's integration settings must be matched; one Laplace point differs from adaptive quadrature.
Stata reference · R reference 
mepoisson count x || id:, intpoints(9)
 
lme4::glmer(count ~ x + (1 | id), data = data, family = stats::poisson(link = "log"), nAGQ = 9, control = lme4::glmerControl(optimizer = "bobyqa"))
lme4 		Poisson random-intercept model with adaptive quadrature
Differs from Stata: Fixed workflow; adapt the named data columns explicitly. Stata's integration settings must be matched; one Laplace point differs from adaptive quadrature.
Stata reference · R reference 
meprobit binary x || id:, intpoints(1)
 
lme4::glmer(binary ~ x + (1 | id), data = data, family = stats::binomial(link = "probit"), nAGQ = 1, control = lme4::glmerControl(optimizer = "bobyqa"))
lme4 		Probit random-intercept model with Laplace approximation
Differs from Stata: Fixed workflow; adapt the named data columns explicitly. Stata's integration settings must be matched; one Laplace point differs from adaptive quadrature.
Stata reference · R reference 
meprobit binary x || id:, intpoints(9)
 
lme4::glmer(binary ~ x + (1 | id), data = data, family = stats::binomial(link = "probit"), nAGQ = 9, control = lme4::glmerControl(optimizer = "bobyqa"))
lme4 		Probit random-intercept model with adaptive quadrature
Differs from Stata: Fixed workflow; adapt the named data columns explicitly. Stata's integration settings must be matched; one Laplace point differs from adaptive quadrature.
Stata reference · R reference 
mixed y c.x##c.x || id:
 
lme4::lmer(y ~ x + I(x^2) + (1 | id), data = data, REML = FALSE)
lme4 		Quadratic fixed trend with random intercept
Differs from Stata: Fixed workflow; adapt the named data columns explicitly. R default REML differs from Stata mixed default ML; explicitly select the target likelihood.
Stata reference · R reference 
mixed y x || id:
 
lme4::lmer(y ~ x + (1 | id), data = data, REML = FALSE)
lme4 		REML random-intercept model
Differs from Stata: REML=FALSE explicitly matches Stata mixed default ML. lme4 does not print Stata large-sample fixed-effect z tests by default; match covariance structures and request the intended inference separately.
Stata reference · R reference 
mixed y x || id: x, covariance(independent)
 
lme4::lmer(y ~ x + (x || id), data = data, REML = FALSE)
lme4 		Independent random intercept and slope
Differs from Stata: REML=FALSE explicitly matches Stata mixed default ML. lme4 does not print Stata large-sample fixed-effect z tests by default; match covariance structures and request the intended inference separately.
Stata reference · R reference 
mixed y x || id: x, covariance(unstructured)
 
lme4::lmer(y ~ x + (x | id), data = data, REML = FALSE)
lme4 		Correlated random intercept and slope
Differs from Stata: REML=FALSE explicitly matches Stata mixed default ML. lme4 does not print Stata large-sample fixed-effect z tests by default; match covariance structures and request the intended inference separately.
Stata reference · R reference 
mixed y x || id:, mle
 
lme4::lmer(y ~ x + (1 | id), data = data, REML = FALSE)
lme4 		ML random-intercept model
Differs from Stata: Fixed workflow; adapt the named data columns explicitly. R default REML differs from Stata mixed default ML; explicitly select the target likelihood.
Stata reference · R reference 
predict fitted, fitted
 
stats::predict(fit)
lme4 		Conditional fitted values including random effects
Differs from Stata: Fixed workflow; adapt the named data columns explicitly. Prediction level is explicit; population and cluster-specific predictions have different interpretations.
Stata reference · R reference 
predict residual, residuals
 
stats::residuals(fit)
lme4 		Conditional mixed-model residuals
Differs from Stata: Fixed workflow; adapt the named data columns explicitly. Prediction level is explicit; population and cluster-specific predictions have different interpretations.
Stata reference · R reference 
predict u0 u1, reffects
 
lme4::ranef(fit)$id
lme4 		Empirical Bayes random intercepts and slopes
Differs from Stata: Fixed workflow; adapt the named data columns explicitly. Prediction level is explicit; population and cluster-specific predictions have different interpretations.
Stata reference · R reference 
predict xb, xb
 
stats::predict(fit, re.form = NA)
lme4 		Fixed-effects-only predictions
Differs from Stata: Fixed workflow; adapt the named data columns explicitly. Prediction level is explicit; population and cluster-specific predictions have different interpretations.
Stata reference · R reference 
quadchk
 
{ fits <- lapply(c(5, 9, 15), function(q) lme4::glmer(binary ~ x + (1 | id), data = data, family = stats::binomial(), nAGQ = q)); data.frame(points = c(5, 9, 15), logLik = vapply(fits, function(f) as.numeric(stats::logLik(f)), numeric(1)), slope = vapply(fits, function(f) unname(lme4::fixef(f)[2]), numeric(1))) }
lme4 		Sensitivity of mixed-logit estimates to quadrature points
Fixed workflow; adapt the named data columns explicitly. Manual quadrature sensitivity check; not an automatic convergence guarantee.
Stata reference · R reference 
xtologit ordinal x
 
ordinal::clmm(ordinal ~ x + (1 | id), data = data, link = "logit", nAGQ = 12)
ordinal 		Random-intercept cumulative-logit model
Differs from Stata: Fixed workflow; adapt the named data columns explicitly. Threshold signs and quadrature defaults must be aligned.
Stata reference · R reference 
xtoprobit ordinal x
 
ordinal::clmm(ordinal ~ x + (1 | id), data = data, link = "probit", nAGQ = 12)
ordinal 		Random-intercept cumulative-probit model
Differs from Stata: Fixed workflow; adapt the named data columns explicitly. Latent residual variance is fixed by the probit link.
Stata reference · R reference

Epidemiology

Stata R Purpose, qualifications and sources 
cc case exposed
 
unname(a * d / (b * c))
base 		Case-control cross-product odds ratio
Differs from Stata: Fixed workflow; adapt the named data columns explicitly. Odds ratio is identified under case-control sampling; absolute disease risk is not.
Stata reference · R reference 
cc case exposed, exact
 
stats::fisher.test(tab, alternative = "two.sided")
stats 		Exact conditional odds-ratio inference
Differs from Stata: Fixed workflow; adapt the named data columns explicitly. Conditional estimate and two-sided p-value conventions can differ.
Stata reference · R reference 
cc case exposed, woolf
 
{ log_or <- log(a * d / (b * c)); se <- sqrt(sum(1 / tab)); exp(c(estimate = log_or, lower = log_or - qnorm(.975) * se, upper = log_or + qnorm(.975) * se)) }
stats 		Approximate log-odds Wald interval
Differs from Stata: Fixed workflow; adapt the named data columns explicitly. Requires nonzero counts; no automatic continuity correction is applied.
Stata reference · R reference 
cci 20 10 80 90
 
unname(a * d / (b * c))
base 		Case-control cross-product odds ratio
Differs from Stata: Fixed workflow; adapt the named data columns explicitly. Odds ratio is identified under case-control sampling; absolute disease risk is not.
Stata reference · R reference 
cci 20 10 80 90, exact
 
stats::fisher.test(tab, alternative = "two.sided")
stats 		Exact conditional odds-ratio inference
Differs from Stata: Fixed workflow; adapt the named data columns explicitly. Conditional estimate and two-sided p-value conventions can differ.
Stata reference · R reference 
cci 20 10 80 90, woolf
 
{ log_or <- log(a * d / (b * c)); se <- sqrt(sum(1 / tab)); exp(c(estimate = log_or, lower = log_or - qnorm(.975) * se, upper = log_or + qnorm(.975) * se)) }
stats 		Approximate log-odds Wald interval
Differs from Stata: Fixed workflow; adapt the named data columns explicitly. Requires nonzero counts; no automatic continuity correction is applied.
Stata reference · R reference 
cs case exposed
 
{ risk1 <- a / (a + c); risk0 <- b / (b + d); c(risk_exposed = risk1, risk_unexposed = risk0, risk_ratio = risk1 / risk0, risk_difference = risk1 - risk0) }
base 		Risk ratio and risk difference
Differs from Stata: Fixed workflow; adapt the named data columns explicitly. Point measures assume a cohort with comparable follow-up; case-control sampling does not identify risk.
Stata reference · R reference 
cs case exposed
 
{ rr <- (a / (a + c)) / (b / (b + d)); af_exposed <- 1 - 1 / rr; risk_population <- (a + b) / (a + b + c + d); c(attributable_fraction_exposed = af_exposed, attributable_fraction_population = (risk_population - b / (b + d)) / risk_population) }
base 		Exposed and population attributable fractions
Differs from Stata: Fixed workflow; adapt the named data columns explicitly. Causal interpretation additionally requires exchangeability and correct study sampling.
Stata reference · R reference 
cs case exposed, exact
 
stats::fisher.test(tab)
stats 		Fisher exact association test for cohort table
Differs from Stata: Fixed workflow; adapt the named data columns explicitly. R reports a conditional maximum-likelihood odds ratio; Stata also reports sample cross-product measures.
Stata reference · R reference 
cs case exposed, or woolf
 
{ odds <- a * d / (b * c); se <- sqrt(1 / a + 1 / b + 1 / c + 1 / d); c(odds_ratio = odds, lower = exp(log(odds) - qnorm(.975) * se), upper = exp(log(odds) + qnorm(.975) * se)) }
stats 		Woolf odds-ratio interval in cohort table
Differs from Stata: Fixed workflow; adapt the named data columns explicitly. Woolf approximation requires positive cell counts.
Stata reference · R reference 
csi 20 10 80 90
 
{ risk1 <- a / (a + c); risk0 <- b / (b + d); c(risk_exposed = risk1, risk_unexposed = risk0, risk_ratio = risk1 / risk0, risk_difference = risk1 - risk0) }
base 		Risk ratio and risk difference
Differs from Stata: Fixed workflow; adapt the named data columns explicitly. Point measures assume a cohort with comparable follow-up; case-control sampling does not identify risk.
Stata reference · R reference 
csi 20 10 80 90
 
{ rr <- (a / (a + c)) / (b / (b + d)); af_exposed <- 1 - 1 / rr; risk_population <- (a + b) / (a + b + c + d); c(attributable_fraction_exposed = af_exposed, attributable_fraction_population = (risk_population - b / (b + d)) / risk_population) }
base 		Exposed and population attributable fractions
Differs from Stata: Fixed workflow; adapt the named data columns explicitly. Causal interpretation additionally requires exchangeability and correct study sampling.
Stata reference · R reference 
csi 20 10 80 90, exact
 
stats::fisher.test(tab)
stats 		Fisher exact association test for cohort table
Differs from Stata: Fixed workflow; adapt the named data columns explicitly. R reports a conditional maximum-likelihood odds ratio; Stata also reports sample cross-product measures.
Stata reference · R reference 
csi 20 10 80 90, or woolf
 
{ odds <- a * d / (b * c); se <- sqrt(1 / a + 1 / b + 1 / c + 1 / d); c(odds_ratio = odds, lower = exp(log(odds) - qnorm(.975) * se), upper = exp(log(odds) + qnorm(.975) * se)) }
stats 		Woolf odds-ratio interval in cohort table
Differs from Stata: Fixed workflow; adapt the named data columns explicitly. Woolf approximation requires positive cell counts.
Stata reference · R reference 
dstdize deaths population age, by(group) base(1)
 
sum(data$deaths / data$population * standard / sum(standard))
base 		Direct standardization of stratum-specific rates
Fixed workflow; adapt the named data columns explicitly. Fixed one-population calculation; standard weights must represent the intended reference population.
Stata reference · R reference 
ir cases exposed time
 
c(rate_exposed = 20 / 100, rate_unexposed = 10 / 200, incidence_rate_ratio = (20 / 100) / (10 / 200), rate_difference = 20 / 100 - 10 / 200)
base 		Incidence-rate ratio and rate difference
Differs from Stata: Fixed workflow; adapt the named data columns explicitly. Counts and person-time denominators, not risks.
Stata reference · R reference 
ir cases exposed time, exact
 
stats::poisson.test(c(20, 10), T = c(100, 200))
stats 		Exact comparison of two Poisson incidence rates
Differs from Stata: Fixed workflow; adapt the named data columns explicitly. Stata ir/iri default mid-p differs; this example requests exact inference.
Stata reference · R reference 
iri 20 10 100 200
 
c(rate_exposed = 20 / 100, rate_unexposed = 10 / 200, incidence_rate_ratio = (20 / 100) / (10 / 200), rate_difference = 20 / 100 - 10 / 200)
base 		Incidence-rate ratio and rate difference
Differs from Stata: Fixed workflow; adapt the named data columns explicitly. Counts and person-time denominators, not risks.
Stata reference · R reference 
iri 20 10 100 200, exact
 
stats::poisson.test(c(20, 10), T = c(100, 200))
stats 		Exact comparison of two Poisson incidence rates
Differs from Stata: Fixed workflow; adapt the named data columns explicitly. Stata ir/iri default mid-p differs; this example requests exact inference.
Stata reference · R reference 
istdize deaths population age using reference, by(group)
 
{ expected <- sum(data$population * reference_rate); observed <- sum(data$deaths); c(observed = observed, expected = expected, SMR = observed / expected) }
base 		Indirect standardization and mortality ratio
Fixed workflow; adapt the named data columns explicitly. Point estimates only; reference strata and person-time units must match.
Stata reference · R reference 
mcc exposed_case exposed_control
 
stats::mcnemar.test(tab, correct = FALSE)
stats 		McNemar comparison of matched binary exposures
Differs from Stata: Fixed workflow; adapt the named data columns explicitly. Uncorrected asymptotic statistic; exact or continuity-corrected alternatives differ.
Stata reference · R reference 
mcc exposed_case exposed_control
 
c(matched_odds_ratio = tab[1, 2] / tab[2, 1], discordant_pairs = tab[1, 2] + tab[2, 1])
base 		Matched odds ratio from discordant pairs
Differs from Stata: Fixed workflow; adapt the named data columns explicitly. Concordant pairs do not inform the matched odds ratio.
Stata reference · R reference 
mcci 30 12 4 54
 
stats::mcnemar.test(tab, correct = FALSE)
stats 		McNemar comparison of matched binary exposures
Differs from Stata: Fixed workflow; adapt the named data columns explicitly. Uncorrected asymptotic statistic; exact or continuity-corrected alternatives differ.
Stata reference · R reference 
mcci 30 12 4 54
 
c(matched_odds_ratio = tab[1, 2] / tab[2, 1], discordant_pairs = tab[1, 2] + tab[2, 1])
base 		Matched odds ratio from discordant pairs
Differs from Stata: Fixed workflow; adapt the named data columns explicitly. Concordant pairs do not inform the matched odds ratio.
Stata reference · R reference 
mhodds case exposed stratum
 
stats::mantelhaen.test(tabs, correct = FALSE)
stats 		Mantel-Haenszel common odds ratio
Differs from Stata: Fixed workflow; adapt the named data columns explicitly. Table orientation determines the odds-ratio direction; assumes a common stratum effect.
Stata reference · R reference 
tabodds case exposed
 
{ out <- stats::aggregate(case ~ exposed, data = data, function(v) c(cases = sum(v), noncases = sum(1 - v))); data.frame(exposed = out$exposed, odds = out$case[, 1] / out$case[, 2]) }
stats 		Odds of disease by exposure category
Fixed workflow; adapt the named data columns explicitly. Computes category odds only; no adjusted trend test.
Stata reference · R reference

Bayesian analysis

Stata R Purpose, qualifications and sources 
bayesmh y x, likelihood(normal({var})) prior({y:}, normal(0,100)) prior({var}, igamma(1,1)) rseed(135)
 
MCMCpack::MCMCregress(y ~ x, data = data, burnin = 100, mcmc = 1000, b0 = 0, B0 = .01, c0 = 2, d0 = 2, seed = 135, verbose = 0)
MCMCpack 		Bayesian Gaussian regression with normal and inverse-gamma priors
Differs from Stata: Fixed workflow; adapt the named data columns explicitly. Simple normal-regression model only; samplers and initialization differ. Short chains demonstrate syntax, not convergence.
Stata reference · R reference 
bayespredict pmean, mean
 
drop(cbind(1, data$x) %*% colMeans(as.matrix(draws)[, c("(Intercept)", "x")]))
MCMCpack 		Posterior mean Gaussian regression response
Differs from Stata: Fixed workflow; adapt the named data columns explicitly. Conditional mean prediction; observation noise is integrated to zero.
Stata reference · R reference 
bayesreps yrep*, nreps(10)
 
{ set.seed(136); keep <- seq(1, nrow(draws), length.out = 10); B <- as.matrix(draws)[keep, ]; means <- cbind(1, data$x) %*% t(B[, c("(Intercept)", "x")]); means + matrix(rnorm(length(means), sd = rep(sqrt(B[, "sigma2"]), each = nrow(data))), nrow(data)) }
MCMCpack 		Replicated outcomes from a Gaussian posterior predictive distribution
Differs from Stata: Fixed workflow; adapt the named data columns explicitly. Includes posterior parameter uncertainty and observation noise; no prediction files are written.
Stata reference · R reference 
bayesstats ess
 
coda::effectiveSize(draws)
coda 		Effective MCMC sample size
Differs from Stata: Fixed workflow; adapt the named data columns explicitly. Spectral-density ESS estimator differs from some Stata autocorrelation/batch calculations.
Stata reference · R reference 
bayesstats summary
 
summary(draws)
coda 		Posterior means, standard deviations and quantiles
Differs from Stata: Fixed workflow; adapt the named data columns explicitly. Summarizes supplied posterior draws; Monte Carlo errors differ by estimator.
Stata reference · R reference 
bayesstats summary (effect:exp({y:x}))
 
summary(exp(draws[, "x", drop = FALSE]))
coda 		Posterior summary of a transformed coefficient
Differs from Stata: Fixed workflow; adapt the named data columns explicitly. Transform draws before summarizing; exp(E[b]) differs from E[exp(b)].
Stata reference · R reference 
bayesstats summary {y:x}, hpd clevel(90)
 
coda::HPDinterval(draws[, "x", drop = FALSE], prob = .9)
coda 		Ninety-percent highest-posterior-density interval
Differs from Stata: Fixed workflow; adapt the named data columns explicitly. HPD algorithms and multimodality handling can differ.
Stata reference · R reference 
bayestest interval {y:x}, lower(0)
 
mean(draws[, "x"] > 0)
base 		Posterior probability of a positive slope
Differs from Stata: Fixed workflow; adapt the named data columns explicitly. Monte Carlo probability of an interval hypothesis; not a frequentist p-value.
Stata reference · R reference 
bayestest interval {y:x}, lower(1.8) upper(2.2)
 
mean(draws[, "x"] > 1.8 & draws[, "x"] < 2.2)
base 		Posterior probability within a substantive effect interval
Differs from Stata: Fixed workflow; adapt the named data columns explicitly. Choose the substantive interval before inspecting posterior results.
Stata reference · R reference

R functions by package

R Package Stata Description
ivreg() AER ivregress Instrumental-Variable Regression
tobit() AER Tobit (censored) regression
dbAppendTable() DBI jdbc Insert rows into a table
dbDisconnect() DBI odbc Disconnect (close) a connection
dbGetQuery() DBI odbc Retrieve results from a query
dbListFields() DBI odbc List field names of a remote table
dbListTables() DBI odbc List remote tables
dbReadTable() DBI jdbc Read database tables as data frames
oblimin() GPArotation rotate Rotations Functions Using Gradient Projection Algorithms
quartimax() GPArotation rotate Rotations Functions Using Gradient Projection Algorithms
boxcox() MASS boxcox Box-Cox Transformations for Linear Models
corresp() MASS ca Simple Correspondence Analysis
glm.nb() MASS Negative binomial regression
isoMDS() MASS mds Kruskal's Non-metric Multidimensional Scaling
lda() MASS discrim Linear Discriminant Analysis
mca() MASS mca Multiple Correspondence Analysis
mvrnorm() MASS drawnorm Simulate from a Multivariate Normal Distribution
polr() MASS Ordered logit or probit
qda() MASS discrim Quadratic Discriminant Analysis
rlm() MASS Robust regression
MCMCregress() MCMCpack bayesmh Markov Chain Monte Carlo for Gaussian Linear Regression
match_data() MatchIt psmatch2 Construct a matched dataset from a matchit object
matchit() MatchIt Matching for causal inference
synth() Synth Synthetic control
SurvS4() VGAM cpoisson Create a Survival Object
binom2.rho() VGAM biprobit Bivariate Probit Regression
cens.poisson() VGAM cpoisson Censored Poisson Family Function
multinomial() VGAM slogit Multinomial Logit Model
posnegbinomial() VGAM tnbreg Positive Negative Binomial Distribution Family Function
posnormal() VGAM truncreg Positive Normal Distribution Family Function
pospoisson() VGAM tpoisson Positive Poisson Distribution Family Function
rrvglm() VGAM slogit Fitting Reduced-Rank Vector Generalized Linear Models (RR-VGLMs) and Doubly Constrained RR-VGLMs (DRR-VGLMs)
vglm() VGAM tpoisson Fitting Vector Generalized Linear Models
weightit() WeightIt Balancing weights, such as inverse-probability weights
.libPaths() base adopath Search Paths for Packages
I() base regress Inhibit Interpretation/Conversion of Objects
Mod() base psdensity Complex Numbers and Basic Functionality
R.home() base sysdir Return the R Home Directory
Sys.Date() base Today's date
Sys.sleep() base sleep Suspend Execution for a Time Interval
Sys.time() base c(current_time) Current date and time
abs() base abs() Absolute value
all() base assert Whether all values are TRUE
any() base count if Whether any value is TRUE
anyDuplicated() base gisid Determine Duplicate Elements
anyNA() base gisid Not Available / Missing Values
apply() base egen rowtotal() Apply a function over rows or columns
as.Date() base Convert to a date
as.POSIXct() base clock Date-time Conversion Functions
as.character() base Convert to text
as.data.frame() base getmata Coerce to a Data Frame
as.double() base Convert to double-precision numbers
as.factor() base encode Convert to a factor
as.integer() base Convert to whole numbers
as.matrix() base cf Matrices
as.numeric() base Convert to numbers
as.vector() base margins Vectors - Creation, Coercion, etc
attr() base Get or set an attribute of an object
attributes() base char Object Attribute Lists
body() base viewsource Access to and Manipulation of the Body of a Function
c() base Combine values into a vector
cat() base Print text without quotes
cbind() base merge 1:1 _n Put side by side (base R)
ceiling() base Round up
charToRaw() base changeeol Convert to or from (Bit/Packed) Raw Vectors
character() base tokenize Character Vectors
chol() base irf The Cholesky Decomposition
class() base describe Class of an object
close() base file Functions to Manipulate Connections (Files, URLs, ...)
col() base spmatrix Column Indexes
colMeans() base margins Form Row and Column Sums and Means
colSums() base missings Form Row and Column Sums and Means
colnames() base ds Column names
crossprod() base tsfilter Matrix Cross-Product
cumprod() base generate Running product
cumsum() base Running sum
cut() base Cut numbers into intervals
data.frame() base range Data Frames
diag() base Diagonal of a matrix
diff() base D. Differences between neighbouring values
dim() base describe Numbers of rows and columns
dir() base List files in a folder
dir.create() base mkdir Create a folder
do.call() base program Call a function with a list of arguments
double() base postfile Double-Precision Vectors
dput() base dataex Write an Object to a File or Recreate it
drop() base lincom Drop Redundant Extent Information
droplevels() base label define Drop unused factor levels
duplicated() base gegen Determine Duplicate Elements
eigen() base varstable Spectral Decomposition of a Matrix
environment() base include Environment Access
exp() base exp() Exponential
expand.grid() base arimasoc Create a Data Frame from All Combinations of Factor Variables
factor() base Categorical variable with labelled levels
file() base file Functions to Manipulate Connections (Files, URLs, ...)
file.copy() base copy File Manipulation
file.exists() base confirm file Whether a file exists
file.info() base hexdump Extract File Information
file.path() base global path Build a file path
floor() base Round down
format() base Format values for display
formatC() base string(x, "%9.2f") Format numbers with width and padding
getwd() base Show the working directory
grep() base unab Pattern Matching and Replacement
grepl() base Whether text matches a pattern
gsub() base subinstr() Replace every match
iconv() base unicode Convert Character Vector between Encodings
identical() base cf Test Objects for Exact Equality
ifelse() base cond() Vectorised if-else (base R)
integer() base gdistinct Integer Vectors
interaction() base egen group() Combine factors into one
invisible() base exit Change the Print Mode to Invisible
is.character() base confirm string variable Whether a vector holds text
is.infinite() base gencohort Finite, Infinite and NaN Numbers
is.na() base Whether values are missing
is.nan() base egen Finite, Infinite and NaN Numbers
is.numeric() base confirm numeric variable Whether a vector holds numbers
lapply() base foreach Apply a function to each element, returning a list
length() base count Number of elements
lengths() base wordcount Lengths of List or Vector Elements
levels() base Categories of a factor
library() base ssc install Attach a package
list() base A list of objects
list.files() base dir Files in a folder
load() base use Load objects from an .RData file
local() base tempvar Evaluate an (Unevaluated) Expression
log() base Logarithm (natural by default)
log10() base log10() Base-10 logarithm
logical() base missings Logical Vectors
ls() base List objects in memory
make.names() base Turn text into valid names
mapply() base arimasoc Apply a Function to Multiple List or Vector Arguments
match() base rename Value Matching
match.arg() base unabcmd Argument Verification Using Partial Matching
matrix() base Create a matrix
max() base Largest value
mean() base Mean
merge() base merge Merge two data frames (base R)
message() base display Print a message
min() base Smallest value
names() base Names of an object, such as column names
nchar() base strlen() Number of characters
ncol() base describe Number of columns
nlevels() base levelsof Number of factor levels
normalizePath() base findfile Express File Paths in Canonical Form
nrow() base Number of rows
numeric() base rangestat Numeric Vectors
options() base set Set global options
order() base sort Positions that would sort a vector
paste() base Join text with a separator
paste0() base Join text without a separator
plot() base Basic plot
pmatch() base unab Partial String Matching
pmax() base max(a, b) Element-wise maximum of several vectors
pmin() base min(a, b) Element-wise minimum of several vectors
print() base Print an object
prop.table() base Turn counts into proportions
proportions() base tabulate, row Turn a table into proportions
qr() base orthog The QR Decomposition of a Matrix
qr.Q() base orthog Reconstruct the Q, R, or X Matrices from a QR Object
range() base summarize Minimum and maximum
rank() base Ranks, with ties averaged
rbind() base append Stack rows (base R)
readBin() base hexdump Transfer Binary Data To and From Connections
readLines() base changeeol Read Text Lines from a Connection
readRDS() base use Read an R object file (.rds)
regexpr() base strpos Pattern Matching and Replacement
rep() base expand Repeat values
rep_len() base Repeat values to a given length
replace() base generate Replace Values in a Vector
replicate() base bootstrap Apply a Function over a List or Vector
requireNamespace() base ftools Loading and Unloading Name Spaces
return() base exit Function Definition
rev() base gsort Reverse the order
rm() base Remove objects from memory
round() base Round to a number of decimals
row() base spmatrix Row Indexes
rowMeans() base egen Form Row and Column Sums and Means
rowSums() base Sum across columns in each row
sample() base sample Random sample of values
sample.int() base bsample Random Samples and Permutations
sapply() base foreach Apply a function to each element and simplify
save() base save Save objects to an .RData file
saveRDS() base Save one R object to a file
scale() base egen std() Standardise columns
scan() base tokenize Read Data Values
seq() base Sequence of numbers
seq_along() base _n Positions of the elements of a vector
seq_len() base _n Whole numbers from 1 to n
serialize() base datasignature Simple Serialization Interface
set.seed() base Make random numbers reproducible
setwd() base Change the working directory
sign() base compare Sign Function
signif() base round() Round to significant digits
sin() base tsfilter Trigonometric Functions
sink() base Send console output to a file
solve() base estat Solve a System of Equations
sort() base Sort a vector
source() base Run an R script
split() base by Split data into groups
sprintf() base Format numbers and text with a template
sqrt() base Square root
stop() base exit 198 Stop with an error
stopifnot() base Stop if a condition is not TRUE
strsplit() base split Split strings
sub() base regexr() Replace the first match
subset() base keep if Rows meeting a condition (base R)
substr() base substr() Part of a string by position
sum() base Sum
summary() base Summary of an object, such as a model
sys.source() base include Parse and Evaluate Expressions from a File
t() base Transpose a matrix
table() base Count values
tapply() base bysort g: egen Apply a function within groups
tempdir() base sysdir Create Names for Temporary Files
tempfile() base Name for a temporary file
tolower() base strlower() Lower case
toupper() base strupper() Upper case
traceback() base Show where the last error happened
transform() base expandcl Transform an Object, for Example a Data Frame
trimws() base strtrim() Remove leading and trailing spaces
trunc() base int() Drop the decimals
try() base capture Run code and continue after an error
tryCatch() base capture Handle errors and warnings
typeof() base describe Storage type
unclass() base rotate Object Classes
unique() base Distinct values
unname() base cf Remove names or dimnames
unsplit() base permute Divide into Groups and Reassemble
vapply() base gdistinct Apply a Function over a List or Vector
vector() base irf Vectors - Creation, Coercion, etc
warning() base display as error Give a warning
which() base list if Positions where a condition is TRUE
with() base Evaluate code with the columns of a data frame
writeBin() base changeeol Transfer Binary Data To and From Connections
writeLines() base checksum Write Lines to a Connection
xor() base egen Logical Operators
betareg() betareg betareg Beta Regression for Rates and Proportions
binsreg() binsreg Binned scatter plot and estimation
boot() boot centile Bootstrap Resampling
boot.ci() boot centile Nonparametric Bootstrap Confidence Intervals
augment() broom predict Add fitted values and residuals to the data
glance() broom One-row summary of model statistics
tidy() broom Coefficient table as a data frame
Anova() car anova Type II and III ANOVA tables
avPlots() car Added-variable plots
deltaMethod() car nlcom Estimate and Standard Error of a Nonlinear Function of Estimated Regression Coefficients
leveneTest() car robvar Levene's Test
linearHypothesis() car test Test linear restrictions on coefficients
vif() car Variance inflation factors
knn() class discrim k-Nearest Neighbour Classification
bal.tab() cobalt pstest Display Balance Statistics in a Table
HPDinterval() coda bayesstats Highest Posterior Density intervals
effectiveSize() coda bayesstats Effective sample size for estimating the mean
fread() data.table import delimited Fast reading of large text files
fwrite() data.table export delimited Fast writing of large text files
winsorize() datawizard Winsorise extreme values
att_gt() did Staggered difference in differences (Callaway-Sant'Anna)
across() dplyr Apply the same calculation to several columns
add_count() dplyr bysort g: generate n = _N Add a column with group counts
all_of() dplyr keep `vars' Select columns named in a character vector
anti_join() dplyr Keep rows of the first table with no match in the second
any_of() dplyr capture keep Select listed columns that exist
arrange() dplyr Sort rows
between() dplyr Whether values lie within a range
bind_cols() dplyr Put tables side by side
bind_rows() dplyr Stack tables on top of each other
c_across() dplyr egen rowtotal() Combine columns inside rowwise()
case_match() dplyr recode Recode values by matching them
case_when() dplyr Several if-else conditions in one step
coalesce() dplyr First non-missing value across vectors
consecutive_id() dplyr tsspell Number runs of identical values
contains() dplyr Select columns whose names contain some text
count() dplyr Count rows in each group
cross_join() dplyr Every combination of rows from two tables
cumall() dplyr sum() Running all(): TRUE until the first FALSE
cume_dist() dplyr cumul Cumulative distribution
cur_group() dplyr by Keys of the current group
cur_group_id() dplyr Number of the current group
dense_rank() dplyr egen group() Ranks without gaps
desc() dplyr Sort in descending order inside arrange()
distinct() dplyr Keep unique rows
ends_with() dplyr Select columns whose names end with some text
everything() dplyr _all Select all columns
filter() dplyr Keep rows that meet a condition
first() dplyr x[1] First value
full_join() dplyr Merge keeping all rows of both tables
glimpse() dplyr describe Compact view of all columns
group_by() dplyr Group rows so later steps work within groups
group_split() dplyr by Split a table into one table per group
group_vars() dplyr by Names of the grouping variables
if_all() dplyr egen rowmin() Whether a condition holds in all of several columns
if_any() dplyr egen anymatch() Whether a condition holds in any of several columns
if_else() dplyr One value if a condition holds, another if not
inner_join() dplyr Merge keeping only matched rows
intersect() dplyr merge, keep(match) Rows in both tables
join_by() dplyr joinby Join keys, including inequality and range joins
lag() dplyr Value from the previous row
last() dplyr x[_N] Last value
last_col() dplyr order The last column
lead() dplyr Value from the next row
left_join() dplyr Merge keeping all rows of the first table
matches() dplyr Select columns whose names match a regular expression
min_rank() dplyr egen rank(), track Ranks with ties given the lowest rank
mutate() dplyr Create or change columns
n() dplyr Number of rows in the current group
n_distinct() dplyr distinct Number of distinct values
na_if() dplyr Turn one value into missing
near() dplyr reldif() Compare numbers with a tolerance
nth() dplyr x[5] Value at a given position
ntile() dplyr xtile Split into n groups of nearly equal size
num_range() dplyr keep x1-x5 Select x1, x2 and so on by number
percent_rank() dplyr egen rank() Rank rescaled to 0-1
pick() dplyr varlist Select columns inside mutate() or summarise()
pull() dplyr Extract one column as a vector
reframe() dplyr collapse Summarise to any number of rows per group
relocate() dplyr Move columns to a new position
rename() dplyr Rename columns (new = old)
rename_with() dplyr Rename columns with a function, such as toupper
right_join() dplyr Merge keeping all rows of the second table
row_number() dplyr Row number
rows_update() dplyr merge, update replace Update rows from another table
rowwise() dplyr egen rowmean() Compute row by row
select() dplyr Keep or drop columns
semi_join() dplyr Keep rows of the first table that have a match in the second
setdiff() dplyr merge, keep(master) Rows in the first table only
slice() dplyr Keep rows by position
slice_head() dplyr bysort g: keep if _n <= 3 First rows of each group
slice_max() dplyr gsort -x; keep in 1/5 Rows with the largest values
slice_min() dplyr sort x; keep in 1/5 Rows with the smallest values
slice_sample() dplyr Random sample of rows
slice_tail() dplyr bysort g: keep if _n > _N - 3 Last rows of each group
starts_with() dplyr Select columns whose names start with some text
summarise() dplyr Reduce each group to one row of statistics
summarize() dplyr collapse Same as summarise()
tally() dplyr count Count rows
transmute() dplyr keep Create columns and keep only those
ungroup() dplyr Remove grouping
union() dplyr append Rows in either table, without duplicates
where() dplyr ds, has(type numeric) Select columns by type, such as where(is.numeric)
cohens_d() effectsize esize Cohen's d and Other Standardized Differences
iv_robust() estimatr ivregress Two-Stage Least Squares Instrumental Variables Regression
coefplot() fixest coefplot Coefficient plot for fixest models
etable() fixest esttab Regression table for fixest models
feglm() fixest xtlogit GLM, such as logit, with fixed effects
fenegbin() fixest xtnbreg Negative binomial model with fixed effects
feols() fixest Linear regression with fixed effects, clustering and IV
fepois() fixest Poisson regression with fixed effects
fitstat() fixest estat firststage Fit statistics, including the first-stage F
sunab() fixest eventstudyinteract Sun and Abraham interactions
flextable() flextable Formatted table for Word or PowerPoint
save_as_docx() flextable Save tables to a Word file
fct_collapse() forcats recode Combine levels into groups
fct_count() forcats tabulate Count each level
fct_drop() forcats label define Drop unused levels
fct_infreq() forcats tabulate x, sort Order levels by frequency
fct_lump_n() forcats recode Lump rare levels into Other
fct_na_value_to_level() forcats tabulate x, missing Make missing values a level
fct_recode() forcats label define, modify Rename factor levels
fct_relevel() forcats ib3.x Move factor levels to the front
fct_reorder() forcats graph bar, sort Order levels by another variable
fct_rev() forcats label define Reverse the order of levels
Arima() forecast forecast Fit ARIMA model to univariate time series
forecast() forecast forecast Objects exported from other packages
fracdiff() fracdiff arfima ML Estimates for Fractionally-Differenced ARIMA (p,d,q) models
boottest() fwildclusterboot boottest Fast wild cluster bootstrap inference
geeglm() geepack xtgee Fit Generalized Estimating Equations (GEE)
aes() ggplot2 Map variables to x, y, colour and other plot features
after_stat() ggplot2 histogram Control aesthetic evaluation
annotate() ggplot2 text() Add one annotation, such as text
coord_cartesian() ggplot2 xscale(range()) Zoom in without dropping data
coord_flip() ggplot2 Swap the x and y axes
coord_polar() ggplot2 Polar coordinates, used for pie charts
facet_grid() ggplot2 by() Panels by two variables
facet_wrap() ggplot2 One panel per group
geom_abline() ggplot2 function Line with a given slope and intercept
geom_area() ggplot2 Area plot
geom_bar() ggplot2 Bar chart of counts
geom_boxplot() ggplot2 Box plot
geom_col() ggplot2 graph bar (asis) Bars with given heights
geom_contour() ggplot2 twoway 2D contours of a 3D surface
geom_density() ggplot2 Density curve
geom_errorbar() ggplot2 Error bars or capped ranges
geom_histogram() ggplot2 Histogram
geom_hline() ggplot2 yline() Horizontal reference line
geom_jitter() ggplot2 scatter, jitter() Jittered points
geom_label() ggplot2 mlabel() Text labels in boxes
geom_line() ggplot2 Lines connecting observations
geom_linerange() ggplot2 twoway Vertical intervals: lines, crossbars & errorbars
geom_point() ggplot2 Points (scatter plot)
geom_pointrange() ggplot2 rcap Points with intervals
geom_qq() ggplot2 Quantile-quantile points
geom_qq_line() ggplot2 Reference line for a quantile-quantile plot
geom_rect() ggplot2 twoway Rectangles
geom_ribbon() ggplot2 Shaded band between two values
geom_segment() ggplot2 twoway Line segments and curves
geom_sf() ggplot2 spmap Map from spatial data
geom_smooth() ggplot2 Fitted line or smooth curve
geom_step() ggplot2 line, connect(stairstep) Step plot
geom_text() ggplot2 Text labels on a plot
geom_tile() ggplot2 heatplot Heat-map tiles
geom_violin() ggplot2 Violin plot
geom_vline() ggplot2 Vertical reference line
ggplot() ggplot2 Start a plot
ggsave() ggplot2 Save a plot to a file
labs() ggplot2 title() Title and axis labels
position_dodge() ggplot2 over() Place bars or points side by side
position_jitter() ggplot2 dotplot Jitter points to avoid overplotting
scale_colour_manual() ggplot2 mcolor() Choose colours by hand
scale_fill_brewer() ggplot2 set scheme ColorBrewer fill colours
scale_x_continuous() ggplot2 xlabel() Axis breaks and labels
scale_x_log10() ggplot2 xscale(log) Logarithmic x axis
scale_y_continuous() ggplot2 scatter Position scales for continuous data (x & y)
scale_y_log10() ggplot2 scatter Position scales for continuous data (x & y)
stat_ecdf() ggplot2 cumul Empirical cumulative distribution plot
stat_summary() ggplot2 Plot a summary, such as the mean, of y at each x
stat_summary_2d() ggplot2 heatplot Summarize a variable within two-dimensional bins
theme() ggplot2 graphregion() Fine-tune the look of a plot
theme_bw() ggplot2 Black-and-white plot theme
theme_classic() ggplot2 set scheme Classic plot theme
theme_minimal() ggplot2 set scheme Minimal plot theme
theme_set() ggplot2 Set the default theme for every plot
xlim() ggplot2 xscale(range()) Limits of the x axis
glmmTMB() glmmTMB menbreg Fit Models with TMB
nbinom2() glmmTMB menbreg Family functions for glmmTMB
cv.glmnet() glmnet lasso Cross-validation for glmnet
pairs() graphics Scatter-plot matrix
plot() graphics quantile The Default Scatterplot Function
stem() graphics Stem-and-leaf display
arrow() grid twoway Describe arrows to add to a line
tbl_cross() gtsummary tabulate x1 x2 Cross table
tbl_regression() gtsummary esttab Regression table
tbl_summary() gtsummary Descriptive Table 1
as_factor() haven decode Turn labelled values into a factor
is_tagged_na() haven mvencode "Tagged" missing values
labelled() haven Attach value labels to a vector
print_labels() haven label list Show value labels
read_dta() haven Read a Stata .dta file
read_sas() haven Read a SAS file
read_sav() haven Read an SPSS .sav file
read_xpt() haven import Read and write SAS transport files
tagged_na() haven replace x = .a Stata-style extended missing values such as .a
write_dta() haven Write a Stata .dta file
write_sav() haven savespss Write an SPSS file
write_xpt() haven export Read and write SAS transport files
zap_labels() haven Remove value labels
zap_missing() haven mvdecode Turn special missing values into NA
adorn_totals() janitor tabulate x1 x2 Add totals to a tabyl
clean_names() janitor rename *, lower Clean and lower-case variable names
get_dupes() janitor duplicates list List duplicated rows
remove_empty() janitor missings dropvars Drop empty rows and columns
tabyl() janitor tabulate Frequency and cross tables
look_for() labelled lookfor Find variables by name or label
set_variable_labels() labelled Set variable labels
to_factor() labelled decode Turn labelled values into a factor
val_labels() labelled label values Get or set value labels
var_label() labelled label variable Get or set a variable label
cfa() lavaan sem Fit Confirmatory Factor Analysis Models
fitMeasures() lavaan estat Fit Measures for a Latent Variable Model
lavResiduals() lavaan estat Residuals
modindices() lavaan estat Modification Indices
sem() lavaan Structural equation model
standardizedSolution() lavaan sem Standardized Solution
VarCorr() lme4 estat sd Variance components of a multilevel model
fixef() lme4 quadchk Extract fixed-effects estimates
glmer() lme4 Multilevel generalised linear model
glmerControl() lme4 melogit Control of Mixed Model Fitting
lmer() lme4 Multilevel linear model
ranef() lme4 predict, reffects Predicted random effects
bptest() lmtest Breusch-Pagan test for heteroskedasticity
coeftest() lmtest Coefficient tests with a chosen variance matrix
grangertest() lmtest vargranger Test for Granger Causality
lrtest() lmtest Likelihood-ratio test
resettest() lmtest RESET test for omitted variables
waldtest() lmtest test Wald test of nested models
as_date() lubridate dofc() Convert to a date
day() lubridate Day of the month
days() lubridate d + 7 A period of days, for date arithmetic
days_in_month() lubridate daysinmonth Get the number of days in the month of a date-time
dmy() lubridate date(s, "DMY") Parse day-month-year text into a date
floor_date() lubridate mofd() Round a date down to a unit, such as the month
hour() lubridate hh() Hour of a date-time
interval() lubridate Time span between two dates
make_date() lubridate Date from year, month and day
make_datetime() lubridate mdyhms() Date-time from its parts
mdy() lubridate date(s, "MDY") Parse month-day-year text into a date
minute() lubridate mm() Minute of a date-time
month() lubridate Month of a date
now() lubridate now() Current date and time
quarter() lubridate quarter() Quarter of a date
time_length() lubridate Length of a time span in chosen units, such as years
today() lubridate today() Today's date
wday() lubridate Day of the week
week() lubridate week() Week of the year
yday() lubridate doy() Day of the year
year() lubridate Year of a date
years() lubridate d + 365 A period of years, for date arithmetic
ymd() lubridate Parse text such as 2024-01-31 into a date
ymd_hms() lubridate Parse text into a date-time
avg_comparisons() marginaleffects margins r.x Average contrasts
avg_predictions() marginaleffects Average predictions, like Stata's margins
avg_slopes() marginaleffects Average marginal effects
comparisons() marginaleffects margins r.x Differences between predictions (contrasts)
hypotheses() marginaleffects Tests of linear and nonlinear hypotheses
plot_predictions() marginaleffects Plot predictions, like marginsplot
plot_slopes() marginaleffects marginsplot Plot marginal effects
predictions() marginaleffects margins Predictions with standard errors
slopes() marginaleffects margins, dydx(x) at() Marginal effects for each observation
cumul() metafor meta Cumulative Meta-Analysis for 'rma' Objects
escalc() metafor meta Calculate Effect Sizes and Outcome Measures
leave1out() metafor meta Leave-One-Out Diagnostics for 'rma' Objects
ranktest() metafor meta Rank Correlation Test for Funnel Plot Asymmetry
regtest() metafor meta Regression Test for Funnel Plot Asymmetry
rma() metafor meta Meta-Analysis via Linear (Mixed-Effects) Models
trimfill() metafor meta Trim and Fill Analysis for 'rma.uni' Objects
D1() mice mi Compare two nested models using D1-statistic
D2() mice mi Compare two nested models using D2-statistic
complete() mice mi Extracts the completed data from a mids object
make.predictorMatrix() mice mi Creates a predictorMatrix argument
mice() mice Multiple imputation by chained equations
pool() mice Combine results across imputations
pool.scalar() mice mi Multiple imputation pooling: univariate version
MIcombine() mitools mi Multiple imputation inference
datasummary() modelsummary tabstat Custom summary-statistics tables
datasummary_skim() modelsummary codebook Quick overview of all variables
modelplot() modelsummary Coefficient plot
modelsummary() modelsummary Regression and summary tables
glht() multcomp contrast General Linear Hypotheses
mcp() multcomp contrast General Linear Hypotheses
dmvnorm() mvtnorm lnmvnormalden Multivariate Normal Density and Random Deviates
pmvnorm() mvtnorm binormal Multivariate Normal Distribution
corAR1() nlme xtregar AR(1) Correlation Structure
gls() nlme hetregress Fit Linear Model Using Generalized Least Squares
lme() nlme xtregar Linear Mixed-Effects Models
nlme() nlme menl Nonlinear Mixed-Effects Models
varExp() nlme hetregress Exponential Variance Function
varIdent() nlme xtgls Constant Variance Function
multinom() nnet Multinomial logit
oaxaca() oaxaca Blinder-Oaxaca decomposition
clmm() ordinal meologit Cumulative Link Mixed Models
coords() pROC roctab Coordinates of a ROC curve
roc() pROC roctab Build a ROC curve
roc.test() pROC roccomp Compare two ROC curves
wrap_plots() patchwork Combine several plots
pdata.frame() plm xtset Declare panel data for plm
pgmm() plm Dynamic panel GMM (Arellano-Bond)
phtest() plm hausman Hausman test
plm() plm Panel models: fixed and random effects
plmtest() plm xttest0 Lagrange FF Multiplier Tests for Panel Models
purtest() plm xtunitroot Unit root tests for panel data
vcovBK() plm xtpcse Beck and Katz Robust Covariance Matrix Estimators
vcovHC() plm xtreg Robust Covariance Matrix Estimators
vcovSCC() plm xtreg Driscoll and Kraay (1998) Robust Covariance Matrix Estimator
polychor() polycor tetrachoric Polychoric Correlation
zeroinfl() pscl zip Zero-inflated Count Data Regression
ICC() psych icc Intraclass Correlations (ICC1, ICC2, ICC3 from Shrout and Fleiss)
KMO() psych estat Find the Kaiser, Meyer, Olkin Measure of Sampling Adequacy
alpha() psych alpha Cronbach's alpha
cohen.kappa() psych kap Find Cohen's kappa and weighted kappa coefficients for correlation of two raters
describe() psych summarize Summary statistics (unlike Stata's describe)
fa() psych factor Factor analysis with more methods
mardia() psych mvtest Calculate univariate or multivariate (Mardia's test) skew and kurtosis for a vector, matrix, or data.frame
principal() psych factor Principal components analysis (PCA)
list_rbind() purrr append Stack a list of data frames
map() purrr foreach Apply a function to each element, returning a list
map2() purrr forvalues Apply a function to pairs of elements
map_chr() purrr foreach Apply a function to each element, returning text
map_dbl() purrr foreach Apply a function to each element, returning numbers
possibly() purrr capture Return a default value instead of an error
reduce() purrr merge Combine elements one after another, such as repeated merges
walk() purrr foreach Apply a function for its side effects, such as saving files
ES.h() pwr power Effect size calculation for proportions
pwr.anova.test() pwr power Power calculations for balanced one-way analysis of variance tests
pwr.f2.test() pwr power Power calculations for the general linear model
pwr.norm.test() pwr power Power calculations for the mean of a normal distribution (known variance)
pwr.p.test() pwr power Power calculations for proportion tests (one sample)
pwr.r.test() pwr power Power calculations for correlation test
pwr.t2n.test() pwr power Power calculations for two samples (different sizes) t-tests of means
rq() quantreg Quantile regression
rdbwdensity() rddensity rdbwdensity Bandwidth Selection for Manipulation Testing
rddensity() rddensity rddensity Manipulation Testing Using Local Polynomial Density Estimation
rdbwselect() rdrobust rdbwselect Bandwidth Selection Procedures for Local Polynomial Regression Discontinuity Estimators
rdrobust() rdrobust Regression discontinuity
parse_number() readr destring, ignore() Extract a number from text such as $1,200
read_csv() readr Read a CSV file
read_delim() readr import delimited Read a file with any delimiter
read_tsv() readr import delimited Read a tab-separated file
write_csv() readr Write a CSV file
excel_sheets() readxl import excel, describe Sheet names of an Excel file
read_excel() readxl Read an Excel sheet
install_github() remotes Install a package from GitHub
find_root() rprojroot setroot Find the root of a directory hierarchy
has_file() rprojroot setroot Is a directory the project root?
selection() sampleSelection Heckman sample-selection model
NeweyWest() sandwich newey Newey-West HAC Covariance Matrix Estimation
vcovBS() sandwich Bootstrap variance matrix
vcovCL() sandwich vce(cluster id) Clustered variance matrix
vcovHC() sandwich vce(robust) Heteroskedasticity-robust variance matrix
vcovJK() sandwich Jackknife variance matrix
st_as_sf() sf spset Convert foreign object to an sf object
as_survey_design() srvyr Declare a survey design
survey_mean() srvyr Survey-weighted mean
survey_prop() srvyr Survey-weighted proportion
survey_quantile() srvyr epctile Survey-weighted quantiles
survey_total() srvyr svy: total Survey-weighted total
AIC() stats Akaike information criterion
BIC() stats Bayesian information criterion
Box.test() stats wntestq #1BoxBox-#1PiercePierce and #1LjungLjung-#1BoxBox Tests
Gamma() stats glm Family Objects for Models
HoltWinters() stats tssmooth Holt-Winters Filtering
IQR() stats egen iqr() Interquartile range
StructTS() stats sspace Fit Structural Time Series
TukeyHSD() stats pwcompare, mcompare(tukey) Tukey multiple comparisons
acf() stats corrgram Autocorrelation function
addmargins() stats tabulate x1 x2 Add totals to a table
aggregate() stats collapse Group statistics (base R)
anova() stats anova ANOVA table, or comparison of nested models
aov() stats Analysis of variance
approx() stats ipolate Interpolation Functions
arima() stats ARIMA time-series model
as.dist() stats mdsmat Distance Matrix Computation
ave() stats stgen Group Averages Over Level Combinations of Factors
bartlett.test() stats oneway Bartlett test of equal variances
binom.test() stats bitest Exact binomial test
binomial() stats Binomial family for logit and probit models
cancor() stats canon Canonical Correlations
ccf() stats xcorr Auto- and Cross- Covariance and -Correlation Function Estimation
chisq.test() stats tabulate x1 x2, chi2 Chi-square test
cmdscale() stats mds Classical (Metric) Multidimensional Scaling
coef() stats Model coefficients
complete.cases() stats egen rowmiss() Rows with no missing values
confint() stats regress Confidence intervals of coefficients
cooks.distance() stats predict Regression Deletion Diagnostics
cor() stats Correlation matrix
cor.test() stats Correlation with a significance test
cov() stats correlate, covariance Covariance
cov2cor() stats estat Correlation, Variance and Covariance (Matrices)
cutree() stats cluster Cut a Tree into Groups of Data
dbeta() stats betaden The Beta Distribution
dbinom() stats binomialp The Binomial Distribution
dcauchy() stats cauchyden The Cauchy Distribution
dchisq() stats chi2den The (non-central) Chi-Squared Distribution
density() stats kdensity Kernel density estimate
dexp() stats exponentialden The Exponential Distribution
df() stats Fden The F Distribution
dfbetas() stats dfbeta Regression Deletion Diagnostics
dffits() stats predict Regression Deletion Diagnostics
dgamma() stats gammaden The Gamma Distribution
dhyper() stats hypergeometricp The Hypergeometric Distribution
dist() stats mds Distance Matrix Computation
dlogis() stats logisticden The Logistic Distribution
dnbinom() stats nbinomialp The Negative Binomial Distribution
dnorm() stats normalden() Normal density
dpois() stats poissonp The Poisson Distribution
dt() stats tden The Student t Distribution
dweibull() stats weibullden The Weibull Distribution
ecdf() stats cumul Empirical cumulative distribution
expand.model.frame() stats vce(cluster g) Retrieve additional variables aligned with a fitted model's observations
factanal() stats Factor analysis
filter() stats tssmooth Linear Filtering on a Time Series
fisher.test() stats Fisher's exact test
fitted() stats Fitted values of a model
ftable() stats table Flat multi-way table
gaussian() stats glm Family Objects for Models
glm() stats Generalised linear model, such as logit or Poisson
hatvalues() stats predict Regression Deletion Diagnostics
hclust() stats cluster Hierarchical Clustering
inverse.gaussian() stats glm Family Objects for Models
kmeans() stats K-means clustering
kruskal.test() stats Kruskal-Wallis test
ks.test() stats Kolmogorov-Smirnov test
lm() stats Linear regression
logLik() stats e(ll) Log-likelihood of a model
lowess() stats lowess Lowess smoother
mad() stats egen mad() Median absolute deviation
manova() stats manova Multivariate Analysis of Variance
mantelhaen.test() stats mhodds Cochran-Mantel-#1HaenszelHaenszel Chi-Squared Test for Count Data
mcnemar.test() stats mcc McNemar test for paired proportions
median() stats Median
model.frame() stats estat Extracting the Model Frame from a Formula or Fit
model.matrix() stats xi Design matrix with dummy variables
na.omit() stats Drop rows with missing values
nls() stats nl Nonlinear least squares
nobs() stats e(N) Number of observations used by a model
offset() stats poisson Include an Offset in a Model Formula
optim() stats ml General-purpose optimisation
p.adjust() stats test, mtest Adjust p-values for multiple tests
pacf() stats pac Auto- and Cross- Covariance and -Correlation Function Estimation
pairwise.t.test() stats pwmean Pairwise t tests with adjusted p-values
pbeta() stats ibeta The Beta Distribution
pbinom() stats binomial The Binomial Distribution
pcauchy() stats cauchy The Cauchy Distribution
pchisq() stats chi2tail() Chi-square probability
pexp() stats exponential The Exponential Distribution
pf() stats F The F Distribution
pgamma() stats gammap The Gamma Distribution
phyper() stats hypergeometric The Hypergeometric Distribution
plogis() stats predictnl The Logistic Distribution
pnbinom() stats nbinomial The Negative Binomial Distribution
pnorm() stats normal() Normal cumulative probability
poisson() stats glm Family Objects for Models
poisson.test() stats stir Exact Poisson tests
poly() stats Polynomial terms for a model
power.prop.test() stats power Power Calculations for Two-Sample Test for Proportions
power.t.test() stats power twomeans Power and sample size for a t test
ppoints() stats quantile Ordinates for Probability Plotting
ppois() stats poisson The Poisson Distribution
prcomp() stats Principal components
predict() stats Predictions from a model
princomp() stats pcamat Principal Components Analysis
promax() stats rotate Rotation Methods for Factor Analysis
prop.test() stats Test of a proportion
prop.trend.test() stats nptrend Test for trend in proportions
pt() stats ttail() t-distribution probability
ptukey() stats tukeyprob The Studentized Range Distribution
pweibull() stats weibull The Weibull Distribution
qbeta() stats invibeta The Beta Distribution
qcauchy() stats invcauchy The Cauchy Distribution
qchisq() stats qchi The (non-central) Chi-Squared Distribution
qexp() stats invexponential The Exponential Distribution
qf() stats invF The F Distribution
qgamma() stats invgammap The Gamma Distribution
qlogis() stats invlogistic The Logistic Distribution
qnorm() stats invnormal() Normal quantile
qqplot() stats qqplot Quantile-Quantile Plots
qt() stats invt The Student t Distribution
qtukey() stats invtukeyprob The Studentized Range Distribution
quantile() stats pctile Quantiles
quasibinomial() stats Binomial family that allows survey weights
quasipoisson() stats svy Family Objects for Models
qweibull() stats invweibull The Weibull Distribution
rbeta() stats rbeta The Beta Distribution
rbinom() stats rbinomial() Binomial random numbers
rcauchy() stats rcauchy The Cauchy Distribution
rchisq() stats rchi2 The (non-central) Chi-Squared Distribution
relevel() stats Choose the base category of a factor
reshape() stats icc Reshape Grouped Data
residuals() stats predict, residuals Model residuals
rexp() stats rexponential The Exponential Distribution
rgamma() stats rgamma The Gamma Distribution
rhyper() stats rhypergeometric The Hypergeometric Distribution
rlogis() stats rlogistic The Logistic Distribution
rnbinom() stats rnbinomial The Negative Binomial Distribution
rnorm() stats Normal random numbers
rpois() stats rpoisson The Poisson Distribution
rstandard() stats predict Regression Deletion Diagnostics
rstudent() stats predict Regression Deletion Diagnostics
rt() stats rt The Student t Distribution
runif() stats Uniform random numbers
runmed() stats tssmooth Running Medians -- Robust Scatter Plot Smoothing
rweibull() stats rweibull The Weibull Distribution
sd() stats Standard deviation
setNames() stats labmask Set the Names in an Object
shapiro.test() stats Shapiro-Wilk normality test
sigma() stats estat Extract Residual Standard Deviation 'Sigma'
spectrum() stats pergram Spectral Density Estimation
step() stats stepwise Stepwise model selection
t.test() stats t test and confidence interval of a mean
ts() stats tsset Time-series object
uniroot() stats ciwidth One Dimensional Root (Zero) Finding
update() stats regress Refit a model with changes
var() stats summarize, detail Variance
var.test() stats sdtest F test of equal variances
varimax() stats rotate Rotation Methods for Factor Analysis
vcov() stats Variance-covariance matrix of a model
weighted.mean() stats summarize [aweight = w] Weighted mean
wilcox.test() stats Wilcoxon rank-sum or signed-rank test
xtabs() stats tabulate x1 x2 Cross-tabulation from a formula
stri_reverse() stringi ustrreverse Reverse Each String
stri_trans_nfc() stringi ustrnormalize Perform or Check For Unicode Normalization
stri_trans_nfkc() stringi ustrnormalize Perform or Check For Unicode Normalization
fixed() stringr Match text literally rather than as a regular expression
regex() stringr Regular expression with options such as ignore_case
str_c() stringr generate s = a + b Join strings
str_count() stringr Count matches in a string
str_detect() stringr Whether a string contains a pattern
str_dup() stringr "-" * 10 Repeat a string
str_ends() stringr substr() Whether a string ends with a pattern
str_extract() stringr Extract the matching text
str_extract_all() stringr ustrregexs() Extract every match
str_flatten() stringr levelsof, separate() Collapse a vector into one string
str_glue() stringr display "`x'" Fill a text template
str_length() stringr Number of characters
str_locate() stringr strpos() Position of a match
str_match() stringr regexs() Extract the groups of a regular expression
str_pad() stringr string(x, "%05.0f") Pad a string to a fixed width
str_remove() stringr subinstr() Remove the first match
str_remove_all() stringr subinstr() Remove every match
str_replace() stringr Replace the first match
str_replace_all() stringr Replace every match
str_split() stringr split Split strings into pieces
str_squish() stringr Trim spaces and collapse repeated spaces
str_starts() stringr substr() Whether a string starts with a pattern
str_sub() stringr Part of a string by position
str_subset() stringr keep if regexm() Keep the strings that match
str_to_lower() stringr Lower case
str_to_sentence() stringr strproper() Sentence case
str_to_title() stringr Title case
str_to_upper() stringr Upper case
str_trim() stringr Remove leading and trailing spaces
str_trunc() stringr abbrev() Shorten strings to a width
str_wrap() stringr display Wrap text into lines
word() stringr Extract words from a sentence
as.svrepdesign() survey svyset Convert a survey design to use replicate weights
cv() survey estat Summary statistics for sample surveys
deff() survey estat Summary statistics for sample surveys
degf() survey svydescribe Contingency tables for survey data
postStratify() survey svyset Post-stratify a survey
rake() survey svyset Raking of replicate weight design
svyby() survey svy Survey statistics on subsets
svychisq() survey svy Contingency tables for survey data
svyciprop() survey svy Confidence intervals for proportions
svydesign() survey svyset Declare a survey design (survey package)
svyglm() survey Survey-weighted regression
svymean() survey svy: mean Survey-weighted mean (survey package)
svyratio() survey svy Ratio estimation
svytable() survey svy: tabulate Survey-weighted table
svytotal() survey svy Summary statistics for sample surveys
Surv() survival Survival-time outcome for a model
basehaz() survival predict Alias for the survfit function
clogit() survival Conditional logit
concordance() survival estat Compute the concordance statistic for data or a model
cox.zph() survival estat Test the Proportional Hazards Assumption of a Cox Regression
coxph() survival Cox proportional hazards model
finegray() survival stcrreg Create data for a Fine-Gray model
frailty() survival stcox Random effects terms
strata() survival Stratum term in a survival model formula
survSplit() survival stsplit Split a survival data set at specified times
survdiff() survival Log-rank test
survfit() survival Kaplan-Meier survival curves
survreg() survival Parametric survival model
add_row() tibble set obs Add rows to a table
as_tibble() tibble Convert to a tibble
rownames_to_column() tibble xpose Turn row names into a column
tibble() tibble Create a data frame (tibble)
tribble() tibble input Type a small table row by row
complete() tidyr Add rows for missing combinations of values
crossing() tidyr cross All combinations, sorted and unique
drop_na() tidyr drop if missing(x) Drop rows with missing values
expand() tidyr fillin All combinations of values, dropping other columns
expand_grid() tidyr cross Table of every combination of given values
fill() tidyr Fill missing values with the previous or next value
full_seq() tidyr tsfill Complete sequence of values
nest() tidyr by Nest rows into list-columns, one row per group
pivot_longer() tidyr Reshape from wide to long
pivot_wider() tidyr Reshape from long to wide
replace_na() tidyr replace x = 0 if missing(x) Replace missing values
separate_longer_delim() tidyr split Split text into rows
separate_wider_delim() tidyr Split one text column into several
separate_wider_position() tidyr substr() Split text into columns by position
uncount() tidyr Repeat each row a given number of times
unite() tidyr egen concat() Join columns into one
unnest() tidyr expand Expand list-columns back into rows
as_character() tidystata decode Alias of decode()
as_dummy() tidystata Dummy variables for every category
as_numeric() tidystata destring Alias of destring()
assert() tidystata Stop and count the rows where a condition fails
browse() tidystata View data in a spreadsheet window
capture() tidystata Return the result or report a caught error and return NULL; quiet suppresses warnings and messages
codebook() tidystata Describe every variable: type, label, counts and values
corr() tidystata Complete-case Pearson correlations
cut_quantile() tidystata xtile Alias of xtile()
decode() tidystata Convert to text, using value labels for labelled data
destring() tidystata Convert to numbers, dropping value labels
ds() tidystata List variable names
duplicates_drop() tidystata Drop duplicated rows, keeping the first
duplicates_list() tidystata List duplicated rows
duplicates_report() tidystata Count duplicated rows
duplicates_tag() tidystata Mark duplicated rows
findit() tidystata Find the R equivalent of a Stata command
fre() tidystata Frequency table with a total row
fre2() tidystata tabulate Two-way cross-tabulation with row and column totals
is_first() tidystata Mark the first row of each group
isid() tidystata Check that variables uniquely identify the rows
misstable() tidystata Missing values per variable, or missing-value patterns
mvdecode() tidystata Turn given codes into missing values
mvencode() tidystata Replace missing values with a code
na_if_value() tidystata mvdecode Alias of mvdecode()
pc() tidystata Percentage of the total
pct() tidystata egen pc() Alias of pc()
print_all() tidystata Print every row and column
pwcorr() tidystata Pairwise correlations with counts and p-values
row_max() tidystata Maximum across variables in each row
row_mean() tidystata Mean across variables in each row
row_median() tidystata Median across variables in each row
row_min() tidystata Minimum across variables in each row
row_miss() tidystata Number of missing values in each row
row_non_miss() tidystata Number of observed values in each row
row_nvals() tidystata Number of distinct values in each row, ignoring missing values
row_sd() tidystata Standard deviation across variables in each row
row_sum() tidystata Sum across variables in each row
row_unique() tidystata egen rownvals() Number of distinct values in each row, counting missing as one
set_seed() tidystata Set the random seed, alone or at the end of a pipe
stata_diff() tidystata Difference from the previous period, respecting time gaps
stata_lag() tidystata Value from an earlier period, respecting time gaps
stata_lead() tidystata Value from a later period, respecting time gaps
stata_merge() tidystata Merge two tables and report matches, like Stata's merge
std() tidystata Standardise to mean 0 and standard deviation 1
summ() tidystata Summary statistics for variables
tab() tidystata Frequency table
tab1() tidystata One frequency table per variable
tab2() tidystata Two-way cross-tabulation, or every pairwise table
tab_gen() tidystata tabulate, generate() Alias of as_dummy()
tabstat() tidystata Chosen statistics for variables, optionally by group
timer() tidystata Time how long code takes
tostring() tidystata Convert to text, keeping codes rather than value labels
value_if_na() tidystata mvencode Alias of mvencode()
variables() tidystata List variables with their types and labels
variables_search() tidystata Find variables by name or label
varlist() tidystata Select variables with Stata wildcards such as x*
xtile() tidystata Quantile groups, following Stata's xtile
md5sum() tools checksum Compute MD5 Checksums
garch() tseries arch Fit GARCH Models to Time Series
as_tsibble() tsibble Declare time-series or panel data
fill_gaps() tsibble Add rows for missing time points
yearmonth() tsibble Monthly period
yearquarter() tsibble Quarterly period
ca.jo() urca vecrank Johansen Procedure for VAR
cajorls() urca vec OLS regression of VECM
ur.df() urca dfuller Augmented-Dickey-Fuller Unit Root Test
ur.ers() urca dfgls Elliott, Rothenberg and Stock Unit Root Test
ur.pp() urca pperron Phillips and Perron Unit Root Test
View() utils browse Open the data viewer
capture.output() utils dataex Send Output to a Character String or File
combn() utils tab2 Generate All Combinations of n Elements, Taken m at a Time
data() utils Load a built-in example dataset
getAnywhere() utils Find where a function is defined
glob2rx() utils Turn a wildcard pattern into a regular expression
head() utils First rows
help() utils Open a help page
help.search() utils Search all installed help pages
install.packages() utils Install a package from CRAN
object.size() utils memory Report the Space Allocated for an Object
packageVersion() utils ftools Package Description
read.csv() utils import delimited Read a CSV file (base R)
read.delim() utils import Data Input
read.fwf() utils infix Read Fixed Width Format Files
read.table() utils infile Data Input
sessionInfo() utils about Collect Information About the Current R Session
str() utils Compact view of an object's structure
tail() utils Last rows
unzip() utils unzipfile Extract or List Zip Archives
update.packages() utils Update installed packages
write.csv() utils export delimited Write a CSV file (base R)
write.table() utils outfile Data Output
write_xlsx() writexl Write an Excel file
zipr() zip zipfile Compress Files into 'zip' Archives
na.locf() zoo stfill Last Observation Carried Forward
rollapply() zoo rolling Apply Rolling Functions