Currently, at least decision_tree seems to be failing due to the keyword "presort" being removed in DecisionTreeClassifier. Downgrading sklearn to earlier versions would fix this, but I'm using more recent version (.23 and up) for other reasons. Is there a way to have finer control over what kwargs are passed? Is this even fixable within hyperopt-sklearn? Or do I have to live with using older versions of scikit-learn?
Currently, at least
decision_tree
seems to be failing due to the keyword "presort" being removed inDecisionTreeClassifier
. Downgrading sklearn to earlier versions would fix this, but I'm using more recent version (.23 and up) for other reasons. Is there a way to have finer control over what kwargs are passed? Is this even fixable within hyperopt-sklearn? Or do I have to live with using older versions of scikit-learn?