On scikit-learn, Training random forest with full depth, testing on different depths

Question

I have a classification problem and I would like to estimate the correct parameters by validation.

from sklearn.ensemble import RandomForestClassifier as RF