Hi,
I had some trouble using kabuki.analyze.post_pred_gen with HDDMRegressor class when append_data=True. I debugged it by replacing
sampled_data = sampled_data.join(data.reset_index(), lsuffix='_sampled')
with
sampled_data = sampled_data.merge(data, left_index=True, right_index=True, suffixes=('_sampled', ''))
in kabuki.analyze._post_pred_generate.
My understanding is that join is a legacy pandas function and merge works better but you might have a more elegant solution for this. Let me know if you'd like me send a pr or if you change the code in this or another way.
Hi,
I had some trouble using
kabuki.analyze.post_pred_genwithHDDMRegressorclass whenappend_data=True. I debugged it by replacingsampled_data = sampled_data.join(data.reset_index(), lsuffix='_sampled')with
sampled_data = sampled_data.merge(data, left_index=True, right_index=True, suffixes=('_sampled', ''))in
kabuki.analyze._post_pred_generate.My understanding is that join is a legacy pandas function and merge works better but you might have a more elegant solution for this. Let me know if you'd like me send a pr or if you change the code in this or another way.