Would it be possible allow experiment data to contain pd.Categorical() columns?
Currently getting" Cannot interpret 'CategoricalDtype(..." error. It would be nice so that when dummies are produced for a categorical variable the numerical ordering of the dummies can be prespecified.