remove unused code

scikit-learn-contrib · Apr 19, 2024 · f6fe948 · f6fe948
1 parent 49ca501
commit f6fe948
Show file tree

Hide file tree

Showing 4 changed files with 0 additions and 18 deletions.
diff --git a/hiclass/LocalClassifierPerNode.py b/hiclass/LocalClassifierPerNode.py
@@ -323,14 +323,6 @@ def predict_proba(self, X):
 
         # normalize probabilities
         level_probability_list = _normalize_probabilities(level_probability_list)
-        """
-        level_probability_list = [
-            np.nan_to_num(
-                level_probabilities / level_probabilities.sum(axis=1, keepdims=True)
-            )
-            for level_probabilities in level_probability_list
-        ]
-        """
 
         # combine probabilities horizontally
         level_probability_list = self._combine_and_reorder(level_probability_list)

diff --git a/hiclass/LocalClassifierPerParentNode.py b/hiclass/LocalClassifierPerParentNode.py
@@ -308,14 +308,6 @@ def _predict_proba_remaining_levels(self, X, y):
 
         # normalize probabilities
         level_probability_list = _normalize_probabilities(level_probability_list)
-        """
-        level_probability_list = [
-            np.nan_to_num(
-                level_probabilities / level_probabilities.sum(axis=1, keepdims=True)
-            )
-            for level_probabilities in level_probability_list
-        ]
-        """
 
         return level_probability_list
 

diff --git a/hiclass/_calibration/Calibrator.py b/hiclass/_calibration/Calibrator.py
@@ -103,7 +103,6 @@ def predict_proba(self, X: np.ndarray):
                 for idx, split in enumerate(score_splits):
                     probabilities[:, idx] = self.calibrators[idx].predict_proba(split)
 
-                # probabilities /= probabilities.sum(axis=1, keepdims=True)
                 probabilities = _normalize_probabilities(probabilities)
 
         else:

diff --git a/hiclass/_calibration/VennAbersCalibrator.py b/hiclass/_calibration/VennAbersCalibrator.py
@@ -304,7 +304,6 @@ def predict_proba(self, scores: np.ndarray):
                     probabilities[:, idx] = self.ivaps[idx].predict_proba(scores)
 
             # normalize
-            # probabilities /= probabilities.sum(axis=1, keepdims=True)
             probabilities = _normalize_probabilities(probabilities)
             return probabilities