Predict_proba in ACP-CrossSampler

[PARODBE]

Hi,

Could It be possible implement a function for predict_proba when you are using acp-crossSampler or another ACP methods? I am doing something like this:

I am using a LDA model. And i know that in model adapter you have build a class for obtaining probabilities but in the acp-crossSampler at the moment, no?

Best Regards,
Pablo