'update'

Author: PowderL

README.rst

@@ -120,6 +120,7 @@ or compute local variable importance based on the accuracy metric ::
   model.fit(train_x, train_y)
   lvig_handler = accuracy_LVIG(model)
   ## compute local variable importance
+  ## 
 	local_variable_importance = lvig_handler.compute_feature_importance(train_x, train_y, partition_feature = partition_feature)
   ## as the accuracy-based LVIG is a model-agnostic method, using other model like xgboost and gradient booting decission tree is applicable
   from sklearn.ensemble import GradientBoostingRegressor