P = Pipeline([('scaler', PowerTransformer()),('selection', RFECV(LogisticRegression())),('classifier', BaggingClassifier(random_state=0))])
K = {'classifier__n_estimators': [10,15,20,25,30]}
gridsearch = GridSearchCV(P, K, verbose=1, cv=5, scoring='neg_mean_squared_error', n_jobs=2).fit(X_train,Y_train) show() method — it has the poof() methodshow() and other new features, but we don’t support python 2.7 anymore, so it will also require an update to your python version to get the latest version of yellowbrick