Running classifiers with cuts on the input arrays (python)

I wonder if it is possible to run a classifier (e.g. RandomForestClassifier, GradientBoostingClassifier) with the call accepting cuts on the input data.

To make the question more clear, let's say that we want to run

gbcl = GradientBoostingClassifier(n_estimators=100,
                                  learning_rate=1.0,
                                  max_depth=5,
                                  random_state=1).fit(X_train, y_train)

but use only the events for which X_train[:,3] > 0.

Is there a way to do this other than creating new input arrays?

from Running classifiers with cuts on the input arrays (python)