Extract From Stata

This Python package provides a command-line tool to extract regression tables and hypothesis-test results from Stata log files. It is equipped to translate results into two types of output: csv and tex.

The tool recognizes the following families of Stata tables:

regressions
equal-means tests
hypothesis tests

Installation

To install this package using pip, type

pip install git+https://github.com/gn0/extract-from-stata

or, alternatively,

git clone https://github.com/gn0/extract-from-stata
pip install ./extract-from-stata

Usage

Basic function

The tool looks for labels in the Stata log file to know which tables to extract. The basic labels are:

@regression
@hypothesis
@equalmeans

These should be put in comments so that Stata does not parse them but shows them in the log. They need to precede the corresponding Stata commands:

// @regression
reg y x

reg y x z

// @hypothesis
lincom x + z

// @equalmeans
ttest x == z

Suppose the Stata log is saved in a file called stata.log. The tool operates on this log file. To translate all regression tables into csv, type:

cat stata.log | extract_from_stata -t regression -c

To translate all regression tables into tex:

cat stata.log | extract_from_stata -t regression -l

To add stars to indicate significance levels:

cat stata.log | extract_from_stata -t regression -l -s

As is evident from these examples, the tool is designed to fit into the Unix pipeline. This means that you can combine it with any other tools that operate within the Unix pipeline to modify the Stata log that extract_from_stata takes as input, and the csv table that it generates as output.

Using modifiers for regression output tables

As the default behavior, the constant and categorical variables are suppressed in the output. They can be brought back by adding the @show_constant and the @show_categoricals labels after the @regression label.

// @regression
// @show_constant
reg y x

// @regression
// @show_categoricals
reg y c.x##i.z

Using identifiers to select which tables to extract

If you want to organize your regression results into separate tables, you can add identifiers to the @regression, @hypothesis, and @equalmeans labels. For example,

// @regression(foo)
reg y x

// @regression(foo)
reg y x, robust

// @regression(bar)
reg y z

// @regression(bar)
reg y z, robust

allows you to extract regressions belonging to foo and bar separately:

cat stata.log | extract_from_stata -t regression -i foo -c
cat stata.log | extract_from_stata -t regression -i bar -c

Author

Gabor Nyeki. Contact information is on http://www.gabornyeki.com/.

License

This package is licensed under the Creative Commons Attribution 4.0 International License: http://creativecommons.org/licenses/by/4.0/.