py-rdpackages 0.0.8

A Pythonic Package for Regression Discontinuity

State of the Art RD Packages For Python

Installation

In terminal window: pip install py_rdpackages In R: install.packages('rdrobust')

Introduction and Use

These packages are a work in progress, but are an attempt to create a wrapper to implement the wonderful RD packages found here (https://sites.google.com/site/rdpackages/rdrobust) which utilize R or Stata, so that they can be used in Python directly.

There are three packages in py_rdpackages:

1. rdplot creates plots of the regression discontinuity with a variety of options.
2. rdrobust does the RD and reports the regression results.
3. rdbwselect selects the optimal bandwidth size.

Requirements

Use of the programs requires all of the following packages in Python:

1. rpy2 for running R in Python
2. matplotlib and seaborn for producing high quality graphics
3. pandas, numpy and statsmodels for data manipulation and dataframe reading

Important: you need to have the original rdrobust installed in R (you can find this in the above link).

Limitations

1. ryp2 produced slow pandas DF to R DF conversions, so I use pd.df.to_csv('temp_file_for_rd.csv') as a solution and then delete that same file after doing the analysis. This should be fixed in future versions.
2. I have not written a full set of graphics options for the RD plots. Future versions will allow you to use all of the classic matplotlib tools, and also turn on/off the vertical line and change the coloring of the scatterplot. Also, I do not incorporate standard errors on the scatterplot bins, which would be easy enough to add. In this version, if you want to do some of those things, you should do them after making a function call.
3. Of course, Python calling R and then converting back to python is not ideal. Some future version should get the original C implementation of the code and just work from there.

Hopefully I will make these better in the future, but for now let me know if you spot any bugs.