Goal: StataPy aims to make it easy for Python programmers to
estimate econometric models using Stata.
Motivation: Stata is a wonderful library of econometric models, but
Stata's programming language is less wonderful.
Approach: I hope to accomplish the above goal with the least amount of code necessary.
StataPy can be installed using pip:
pip install -e git+https://github.com/amarder/StataPy.git#egg=StataPy
Here is a simple example of how to estimate a Stata model:
import pandas
import stata
data = pandas.DataFrame({'y': [1, 2, 3, 4], 'x': [0, 1, 0, 1]})
model = stata.Model(data, 'regress y x, robust')
print model.estimate()
By default, stata.Model assumes that Stata can be called on your
computer with the command stata. If the path on your computer is
different, the correct path can be passed into the Model constructor
through the parameter path_to_stata_binary.
Note the example uses a pandas.DataFrame. This is most likely how I
will use StataPy but it is not a requirement. One could use a
stata.Model with any data object that has a to_csv method like a
pandas DataFrame.
The format parameter in the stata.Model constructor is a hack. I'm
not sure how to prevent rounding and also have decimal numbers between
zero and one have a leading zero.
It would be nice to handle temporary files more gracefully.
It would be nice to package this for an easy pip install from github.