def sym_predict_proba_bagging_classifier(estimator):
inputs = syms(estimator)
Var = VariableFactory(existing=inputs)
vars_ = tuple(Var() for _ in range(len(estimator.estimators_)))
calls = tuple(starmap(lambda var, est, args: ((var,), (sym_predict_proba(est) if hasattr(est, 'predict_proba') else sym_predict(est), tuple(map(curry(__getitem__)(inputs), list(args))))),
zip(vars_, estimator.estimators_, estimator.estimators_features_)))
outputs = (reduce(__add__, vars_) / RealNumber(len(estimator.estimators_)),)
return Function(inputs, calls, outputs)
def syms__calibrated_classifier(estimator):
return syms(estimator.base_estimator)
def syms_calibrated_classifier_cv(estimator):
return syms(estimator.calibrated_classifiers_[0])
def syms_pipeline(estimator):
return syms(estimator.steps[0][1])
def syms(self):
return syms(self.estimator_)
def syms_super_learner(estimator):
return syms(first(estimator.cross_validating_estimators_.values()))