def cv_max_features(clf, data, max_features, predictors, target, cv=5):
mses = []
n_features = range(1, max_features)
for n_feature in n_features:
clf.max_features = n_feature
mses.append(cv_mse(clf, data, predictors, target, cv))
return pd.DataFrame({'mse': mses}, index=n_features)
def cv_n_estimators(clf, data, max_estimators, predictors, target,
cv=5, step=1):
mses = []
n_estimators = range(1, max_estimators, step)
for n_estimator in n_estimators:
clf.n_estimators = n_estimator
mses.append(cv_mse(clf, data, predictors, target, cv))
return pd.DataFrame({'mse': mses}, index=n_estimators)