statsmodels.sandbox.regression.try_catdata.groupstatsbin

statsmodels.sandbox.regression.try_catdata.groupstatsbin(factors, values)[source]

uses np.bincount, assumes factors/labels are integers