Is your feature request related to a problem? Please describe.
The CalibratedExplanations currently provide functionality for a mix between a set of explanations and individual explanations.
Describe the solution you'd like
I would like to separate between the set of CalibratedExplanations and individual CalibratedExplanation in a better way.
Describe alternatives you've considered
It would make sense to create subclasses for the CalibratedExplanation representing factual and counterfactual explanations, as these clearly have class specific behaviour. It would also make sense to create subclasses representing regression vs probabilistic explanations, as they differ in regards to the plots. However, would it make sense to create e.g. FactualRegression, CounterfactualRegression, FactualProbabilistic and CounterfactualProbabilistic?
Is your feature request related to a problem? Please describe. The
CalibratedExplanations
currently provide functionality for a mix between a set of explanations and individual explanations.Describe the solution you'd like I would like to separate between the set of
CalibratedExplanations
and individualCalibratedExplanation
in a better way.Describe alternatives you've considered It would make sense to create subclasses for the
CalibratedExplanation
representing factual and counterfactual explanations, as these clearly have class specific behaviour. It would also make sense to create subclasses representing regression vs probabilistic explanations, as they differ in regards to the plots. However, would it make sense to create e.g.FactualRegression
,CounterfactualRegression
,FactualProbabilistic
andCounterfactualProbabilistic
?