brycewang-stanford/Auto-Empirical-Research-Skills / skills/14-luischanci-claude-code-research-starter/dot-claude/skills/stata
stata
VerifiedComprehensive Stata reference for writing correct .do files, data management, econometrics, causal inference, graphics, Mata programming, and 17+ community packages (reghdfe, estout, did, rdrobust, etc.). Covers syntax, options, gotchas, and idiomatic patterns. Use this skill whenever the user asks you to write, debug, or explain Stata code.
packages/tabout.md
View on GitHub# tabout: Publication-Quality Cross-Tabulations and Summary Tables
## Overview
`tabout` extends Stata's built-in `tabulate` with extensive formatting and export capabilities for publication-quality tables.
**Key features:**
- One-way and two-way frequency tables with summary statistics
- Export to LaTeX, HTML, plain text, CSV, and Excel
- Multiple panel tables (append mode)
- Support for survey data and weights (`svy`)
- Cell contents: freq, row/col/cell percentages, mean, sd, median, quartiles, se, min, max, confidence intervals
---
## Installation
```stata
ssc install tabout
help tabout
which tabout
```
---
## Quick Start
```stata
sysuse auto, clear
* Basic one-way frequency table
tabout foreign using table1.txt, replace
* One-way with percentages
tabout foreign using table1.txt, c(freq col) f(0 1) replace
* Basic two-way table with row and column percentages
tabout foreign rep78 using table2.txt, ///
c(freq row col) f(0 1 1) replace
* Export to LaTeX
tabout foreign rep78 using table2.tex, ///
c(freq row col) f(0 1 1) style(tex) replace
```
---
## One-Way Tables
```stata
* Frequencies with percentages and cumulative
tabout foreign using oneway1.txt, ///
c(freq col cum) f(0 1 1) replace
* With custom column labels and totals
tabout foreign using oneway2.txt, ///
c(freq col) f(0 1) ///
clab(N %) ///
ptotal(both) ///
replace
* Stack multiple one-way tables
tabout foreign rep78 using oneway3.txt, ///
oneway ///
c(freq col) f(0 1) ///
clab(N %) ///
replace
* One-way summary statistics
tabout foreign rep78 using oneway3.txt, ///
oneway ///
sum ///
c(mean price sd price) ///
f(0 1) ///
clab(Mean SD) ///
replace
```
---
## Two-Way Tables
### Cell Contents Options
The `c()` option controls what appears in cells. Each statistic needs a corresponding decimal format in `f()`.
```stata
* Frequencies, row and column percentages
tabout foreign rep78 using twoway.txt, ///
c(freq row col) f(0 1 1) ///
clab(N Row% Col%) ///
replace
* Cell percentages (of total)
tabout foreign rep78 using twoway.txt, ///
c(cell) f(1) replace
* All four in one table
tabout foreign rep78 using twoway.txt, ///
c(freq row col cell) f(0 1 1 1) ///
clab(N Row% Col% Cell%) ///
replace
```
### Totals and Tests
```stata
* Row and column totals with chi-square test
tabout foreign rep78 using twoway.txt, ///
c(freq row) f(0 1) ///
clab(N Row%) ///
ptotal(both) ///
stats(chi2) ///
replace
```
**`ptotal()` values:** `single` (column total), `both` (row + column), `none`
---
## Summary Statistics by Groups
```stata
sysuse auto, clear
* Mean, SD, Min, Max by group
tabout foreign using summary.txt, ///
sum ///
c(mean price sd price min price max price) ///
f(2 2 2 2) ///
clab(Mean SD Min Max) ///
replace
* Median and quartiles
tabout foreign using summary.txt, ///
sum ///
c(median price q1 price q3 price) ///
f(2 2 2) ///
clab(Median Q1 Q3) ///
replace
* Mean with SE and confidence intervals
tabout foreign using summary.txt, ///
sum ///
c(mean price se price lb price ub price) f(2 2 2 2) ///
clab(Mean SE Lower Upper) ///
replace
* Two-way summary: Mean and SD
tabout foreign rep78 using summary2.txt, ///
sum ///
c(mean price sd price) f(2 2) ///
clab(Mean SD) ///
replace
* Multiple variables summary (stacked)
tabout foreign using summary3.txt, ///
sum ///
c(mean price sd price) ///
c(mean mpg sd mpg) ///
c(mean weight sd weight) ///
f(2 2) ///
clab(Mean SD) ///
replace
```
---
## Multiple Panel Tables
Build panels using `replace` for the first, then `append` for subsequent panels. Use `h1()` / `h2()` for panel headings.
```stata
* Panel A
tabout foreign using panel.txt, ///
c(freq col) f(0 1) ///
h1(Panel A: Origin) ///
ptotal(single) ///
replace
* Panel B (appended)
tabout rep78 using panel.txt, ///
c(freq col) f(0 1) ///
h1(Panel B: Repair Record) ///
ptotal(single) ///
append
* Panel C: Summary stats panel (appended)
tabout foreign using panel.txt, ///
sum ///
c(mean price sd price) f(2 2) ///
clab(Mean SD) ///
h1(Panel C: Price by Origin) ///
append
```
### Vertical Panels (Layout 2)
```stata
tabout foreign using panel2.txt, ///
c(freq col) f(0 1) ///
h2(Domestic vs Foreign) ///
layout(2) ///
replace
```
---
## Output Formats
### Plain Text / CSV
```stata
* Tab-delimited (Excel-compatible)
tabout foreign rep78 using table.txt, ///
c(freq row) f(0 1) ///
delimit(tab) replace
* CSV
tabout foreign rep78 using table.csv, ///
c(freq row) f(0 1) ///
delimit(",") replace
```
### LaTeX
```stata
* LaTeX with booktabs, caption, and label
tabout foreign rep78 using table.tex, ///
c(freq row col) f(0 1 1) ///
clab(N Row\% Col\%) ///
style(tex) bt ///
caption(Cross-tabulation of car origin and repair record) ///
lab(tab:cross) ///
ptotal(both) ///
replace
```
**Gotcha:** Escape special characters in LaTeX: `\%`, `\$`, `\_`, `\&`, `\#`
### HTML
```stata
tabout foreign rep78 using table.html, ///
c(freq row) f(0 1) ///
style(html) ///
title(Table 1: Car Origin by Repair Record) ///
fn(Source: 1978 Automobile Data) ///
replace
```
---
## Survey Data and Weights
```stata
* Setup survey design first
webuse nhanes2, clear
svyset psuid [pweight=finalwgt], strata(stratid)
* Survey-weighted frequency table
tabout sex race using survey.txt, ///
c(freq row) f(0 1) ///
svy ///
replace
* Survey-weighted summary statistics
tabout sex race using survey.txt, ///
sum ///
c(mean bpsystol se bpsystol) f(2 2) ///
clab(Mean SE) ///
svy ///
replace
* Subpopulation analysis (proper survey subpop)
tabout sex race using survey.txt, ///
sum ///
c(mean bpsystol) f(2) ///
svy ///
subpop(if age >= 40) ///
replace
```
**Gotcha:** Use `subpop()` for proper survey subpopulation estimation rather than `if` conditions, which can produce incorrect standard errors.
---
## Customization Options
### Decimal Places
`f()` takes one value per statistic in `c()`:
```stata
* 0 decimals for freq, 1 for row%, 1 for col%
tabout foreign rep78 using t.txt, c(freq row col) f(0 1 1) replace
```
### Column Labels
```stata
* Underscores become spaces in clab()
tabout foreign using t.txt, ///
sum c(mean price sd price) f(2 2) ///
clab(Mean_Price SD_Price) replace
```
### Headings
```stata
tabout foreign rep78 using t.txt, ///
c(freq row) f(0 1) ///
h1(Table 1: Descriptive Statistics) ///
h2(Panel A: Cross-tabulation) ///
h3(By Origin and Repair Record) ///
replace
```
### Layout Options
| Layout | Description |
|--------|-------------|
| `layout(1)` | Compact (default) |
| `layout(2)` | Column layout |
| `layout(3)` | Wide layout |
### Missing Values
```stata
* Include missing values as a category
tabout foreign rep78 using t.txt, ///
c(freq row) f(0 1) mi replace
* Or explicitly exclude
tabout foreign rep78 if !missing(foreign, rep78) using t.txt, ///
c(freq row) f(0 1) replace
```
---
## Template: Baseline Characteristics Table (Table 1)
```stata
sysuse auto, clear
generate treatment = foreign
label define trt 0 "Control" 1 "Treatment"
label values treatment trt
* Panel A: Categorical variables
tabout treatment using table1.tex, ///
c(freq col) f(0 1) ///
clab(N \%) ///
h1(Panel A: Categorical Variables) ///
ptotal(single) ///
style(tex) bt replace
tabout rep78 treatment using table1.tex, ///
c(freq col) f(0 1) ///
clab(N \%) ///
ptotal(both) stats(chi2) ///
style(tex) bt append
* Panel B: Continuous variables (loop)
local vars "price mpg weight"
local labels `" "Price (\$)" "MPG" "Weight (lbs)" "'
local i = 1
foreach var of local vars {
local lab : word `i' of `labels'
tabout treatment using table1.tex, ///
sum ///
c(N `var' mean `var' sd `var') f(0 2 2) ///
clab(N Mean SD) ///
h2(\textit{`lab'}) ///
style(tex) bt append
local ++i
}
```
---
## Template: Multi-Format Output Program
```stata
capture program drop mytable
program define mytable
syntax using/, [format(string)]
if "`format'" == "tex" {
local style "style(tex) bt"
local pct "\%"
}
else if "`format'" == "html" {
local style "style(html)"
local pct "%"
}
else {
local style ""
local pct "%"
}
tabout foreign rep78 `using', ///
c(freq row) f(0 1) ///
clab(N Row`pct') ///
`style' ptotal(both) replace
end
mytable using table.txt
mytable using table.tex, format(tex)
mytable using table.html, format(html)
```
---
## Template: Batch Stratified Tables
```stata
sysuse auto, clear
egen price_strata = cut(price), group(3)
replace price_strata = price_strata + 1
label define stratalab 1 "Low Price" 2 "Medium Price" 3 "High Price"
label values price_strata stratalab
* Combined file: overall + each stratum
tabout foreign rep78 using combined.tex, ///
c(freq row col) f(0 1 1) ///
h1(Table 1: All Strata Combined) ///
ptotal(both) style(tex) bt replace
foreach s in 1 2 3 {
local slab : label stratalab `s'
tabout foreign rep78 if price_strata == `s' using combined.tex, ///
c(freq row col) f(0 1 1) ///
h1(Table `=`s'+1': `slab' Stratum) ///
ptotal(both) style(tex) bt append
}
```
---
## Advanced Tips
### Conditional Table Generation
```stata
count if foreign == 1
if r(N) >= 20 {
tabout foreign rep78 using conditional.tex, ///
c(freq row) f(0 1) style(tex) bt replace
}
else {
display as error "Insufficient sample size"
}
```
### Dynamic Column Labels
```stata
local today : display %tdCY-N-D date(c(current_date), "DMY")
local clab "N_as_of_`today' Percent"
tabout foreign using dynamic.txt, ///
c(freq col) f(0 1) clab(`clab') replace
```
### Locals for Consistent Formatting
```stata
local fmt_freq "0"
local fmt_pct "1"
local fmt_mean "2"
tabout foreign using t.txt, ///
sum c(mean price sd price) ///
f(`fmt_mean' `fmt_mean') replace
```
---
## Common Issues
| Issue | Solution |
|-------|----------|
| LaTeX special chars (`%`, `$`, `_`, `&`, `#`) | Escape with backslash: `\%`, `\$`, etc. |
| Missing values in table | Use `mi` option, or filter with `if !missing()` |
| Long variable labels in headers | Shorten labels or use `h2()` / `h3()` to override |
| Alignment problems | Use `style(txt)` for text; `style(tex) bt` for LaTeX booktabs |
| Survey weights not applied | Run `svyset` first, then use `svy` option |
---
## Quick Reference
### Key Options
| Option | Description |
|--------|-------------|
| `c()` | Cell contents: `freq`, `row`, `col`, `cell`, `cum`, `mean`, `sd`, `median`, `q1`, `q3`, `min`, `max`, `se`, `lb`, `ub`, `N` |
| `f()` | Decimal places for each statistic in `c()` |
| `clab()` | Column labels (underscores become spaces) |
| `h1()`, `h2()`, `h3()` | Hierarchical headings |
| `style()` | Output: `tex`, `html`, `tab`, `csv` |
| `bt` | LaTeX booktabs formatting |
| `ptotal()` | Totals: `single`, `both`, `none` |
| `sum` | Summary statistics mode |
| `svy` | Use survey weights (requires prior `svyset`) |
| `mi` | Include missing values |
| `oneway` | Stack multiple variables as one-way tables |
| `layout()` | Layout: `1` (compact), `2` (column), `3` (wide) |
| `replace` / `append` | File write mode |
| `stats()` | Statistical tests: `chi2` |
| `delimit()` | Delimiter: `tab`, `","` |
| `caption()` | LaTeX table caption |
| `lab()` | LaTeX table label |
| `title()` | HTML title |
| `fn()` | Footnote text |
| `subpop()` | Survey subpopulation |
### Common Patterns
```stata
* Basic frequency: c(freq) f(0)
* Freq + row%: c(freq row) f(0 1)
* Freq + row% + col%: c(freq row col) f(0 1 1)
* Summary stats: sum c(mean var sd var) f(2 2)
* Full summary: sum c(N var mean var sd var median var) f(0 2 2 2)
* LaTeX with booktabs: style(tex) bt
* Survey-weighted: svy (after svyset)
* Multi-panel: replace on first, append on rest
```
### Related Packages
```stata
ssc install estout // Regression tables
ssc install fre // Enhanced frequency tables
ssc install table1 // Medical-style baseline tables
```
Related skills
| Skill | Stars | Updated | Works with | Status | Install |
|---|---|---|---|---|---|
| finishing-a-development-branchobra/superpowersUse when implementation is complete, all tests pass, and you need to decide how to integrate the work | Stars: 323k | Updated | Claude CodeCodex | Verified | Install |
| systematic-debuggingobra/superpowersUse when encountering any bug, test failure, or unexpected behavior, before proposing fixes | Stars: 323k | Updated | Claude CodeCodex | Verified | Install |
| test-driven-developmentobra/superpowersUse when implementing any feature or bugfix, before writing implementation code | Stars: 323k | Updated | Claude CodeCodex | Verified | Install |
| skill-creatoranthropics/skillsCreate new skills, modify and improve existing skills, and measure skill performance. Use when users want to create a skill from scratch, edit, or optimize an existing skill, run evals to test a skill, benchmark skill performance with variance analysis, or optimize a skill's description for better triggering accuracy. | Stars: 198k | Updated | Claude CodeCodex | Verified | Install |