Was trying to replicate the plots as mentioned in the paper, but no exact results. I think this is to do with slightly different matplotlib parameters and different feature_intervals & feature_di used in the example notebook.
If I can get some help here with the settings?

