rdd

rdd copied to clipboard

Predictions on new data X are not yet intuitive, when using the native OLS.predict() function. The DataFrame X one wants to predict requires a column *TREATED*. I suggest to add...

ojwenzel

I'm running the following line in a jupyter notebook to compute the optimal bandwidth. `bandwidth_opt = rdd.optimal_bandwidth(data[predictor], data[target], cut=threshold)` where data is a pandas data frame. The columns have 20...

ojwenzel

add interaction term to rdd function

1

Thanks for the package. A request for a minor change, I propose adding an argument to the RDD function to allow for an interaction term between the treatment and running...

darrelrobinson

Wrote functions for visual balance checks and balance checks using placebo outcomes

divyansha

- Add error checking in functions - Robustly test functions for input errors, mismatching sizes, etc - Don't only through errors when one occurs, try to solve the issues I...

evan-magnusson

Remove the double for loop in the `bin_data()` function, if possible. Potential methods: - groupby - pd.cut() - pd.where()

evan-magnusson

In the `optimal_bandwidth()` function, allow the user to specify alternative kernels.

evan-magnusson

In the `optimal_bandwidth()` function, the optimal bandwidth is computed as if no controls will be in the estimation equation. Research and implement how to alter this function to allow for...

evan-magnusson