statsmodels.sandbox.stats.multicomp.TukeyHSDResults.summary_frame¶

TukeyHSDResults.summary_frame()[source]¶

Summary DataFrame

The group columns are labeled as “group_t” and “group_c” with mean difference defined as treatment minus control. This should be less confusing than numeric labels group1 and group2.

Returns:¶

pandas.DataFrame

Notes

The number of columns will likely increase in a future version of statsmodels. Do not use numeric indices for the DataFrame in order to be robust to the addition of columns.