Merge pull request #20 from loft-br/feat/add-feature-importance

feat: make feature importance transparent

Author: GabrielGimenez

xgbse/_debiased_bce.py

@@ -149,6 +149,7 @@ def __init__(
         self.lr_params = lr_params
         self.n_jobs = n_jobs
         self.persist_train = False
+        self.feature_importances_ = None
 
     def fit(
         self,
@@ -224,7 +225,7 @@ def fit(
             evals=evals,
             verbose_eval=verbose_eval,
         )
-
+        self.feature_importances_ = self.bst.get_score()
         # predicting and encoding leaves
         self.encoder = OneHotEncoder()
         leaves = self.bst.predict(dtrain, pred_leaf=True)

xgbse/_kaplan_neighbors.py

@@ -98,6 +98,7 @@ def __init__(self, xgb_params=None, n_neighbors=30, radius=None):
         self.persist_train = False
         self.index_id = None
         self.radius = None
+        self.feature_importances_ = None
 
     def fit(
         self,
@@ -171,6 +172,7 @@ def fit(
             evals=evals,
             verbose_eval=verbose_eval,
         )
+        self.feature_importances_ = self.bst.get_score()
 
         # creating nearest neighbor index
         leaves = self.bst.predict(dtrain, pred_leaf=True)
@@ -338,6 +340,7 @@ def __init__(
         self.xgb_params = xgb_params
         self.persist_train = False
         self.index_id = None
+        self.feature_importances_ = None
 
     def fit(
         self,
@@ -388,6 +391,7 @@ def fit(
 
         # training XGB
         self.bst = xgb.train(self.xgb_params, dtrain, num_boost_round=1, **xgb_kwargs)
+        self.feature_importances_ = self.bst.get_score()
 
         # getting leaves
         leaves = self.bst.predict(dtrain, pred_leaf=True)

xgbse/_stacked_weibull.py

@@ -94,6 +94,7 @@ def __init__(
         self.xgb_params = xgb_params
         self.weibull_params = weibull_params
         self.persist_train = False
+        self.feature_importances_ = None
 
     def fit(
         self,
@@ -167,6 +168,7 @@ def fit(
             evals=evals,
             verbose_eval=verbose_eval,
         )
+        self.feature_importances_ = self.bst.get_score()
 
         # predicting risk from XGBoost
         train_risk = self.bst.predict(dtrain)