Clustered regression

1. Setup your workspace:

   from __future__ import print_function
   from __future__ import absolute_import
   from __future__ import division
   
   import pandas as pd
   import numpy as np
   
   from py_metrics import caches
   from py_metrics.regress import Cluster
   
   frame = pd.read_csv(caches.data_path('ddk2011.txt'))
   frame['intercept'] = 1.0
   std = frame['totalscore'].std()
   mu = frame['totalscore'].mean()
   frame['testscore']  = (frame['totalscore'] - mu) / std

2. Fit the regression:

   # Initialize
   x = ['intercept', 'tracking']
   y = 'testscore'
   grp = 'schoolid'
   
   reg = Cluster(x, y, grp)
   reg.fit(frame)
   reg.summarize()

3. Estimate a cluster-robust covariance matrix:

   vce = pd.DataFrame(
       reg.vce('cr3'),
       index=reg.x_cols,
       columns=reg.x_cols)
   print(vce)

Additional examples with more detail are available in the examples directory.