Merge pull request scikit-learn-contrib#34 from tagomatech/tagomatech-patch-3

replaced parameter value in example code

Author: danielhomola

boruta/boruta_py.py

@@ -150,7 +150,7 @@ class BorutaPy(BaseEstimator, TransformerMixin):
     
     # define random forest classifier, with utilising all cores and
     # sampling in proportion to y labels
-    rf = RandomForestClassifier(n_jobs=-1, class_weight='auto', max_depth=5)
+    rf = RandomForestClassifier(n_jobs=-1, class_weight='balanced', max_depth=5)
     
     # define Boruta feature selection method
     feat_selector = BorutaPy(rf, n_estimators='auto', verbose=2, random_state=1)