review

Author: GitName

Diff for: doc/metrics.rst

@@ -45,6 +45,18 @@ The :func:`make_index_balanced_accuracy` :cite:`garcia2012effectiveness` can
 wrap any metric and give more importance to a specific class using the
 parameter ``alpha``.
 
+.. _macro_averaged_mean_absolute_error:
+
+Macro-Averaged Mean Absolute Error (MA-MAE)
+-------------------------------------------
+
+Ordinal classification is used when there is a rank among classes, for example
+levels of functionality or movie ratings.
+
+The :func:`macro_averaged_mean_absolute_error` :cite:`esuli2009ordinal` is used
+for imbalanced ordinal classification. We compute each MAE for each class and
+average them, giving an equal weight to each class.
+
 .. _classification_report:
 
 Summary of important metrics
@@ -54,10 +66,3 @@ The :func:`classification_report_imbalanced` will compute a set of metrics
 per class and summarize it in a table. The parameter `output_dict` allows
 to get a string or a Python dictionary. This dictionary can be reused to create
 a Pandas dataframe for instance.
-
-.. _macro_averaged_mean_absolute_error:
-
-Macro-Averaged Mean Absolute Error (MA-MAE)
--------------------------------------------
-
-The :func:`macro_averaged_mean_absolute_error` :cite:`esuli2009ordinal` is used for imbalanced ordinal classification. We compute each MAE for each class and average them, giving an equal weight to each class.

Diff for: imblearn/metrics/_classification.py

@@ -1012,10 +1012,10 @@ def macro_averaged_mean_absolute_error(y_true, y_pred):
 
     Parameters
     ----------
-    y_true : 1d array-like, or label indicator array / sparse matrix
+    y_true : array-like of shape (n_samples,) or (n_samples, n_outputs)
         Ground truth (correct) target values.
 
-    y_pred : 1d array-like, or label indicator array / sparse matrix
+    y_pred : array-like of shape (n_samples,) or (n_samples, n_outputs)
         Estimated targets as returned by a classifier.
 
     Returns
@@ -1033,20 +1033,19 @@ def macro_averaged_mean_absolute_error(y_true, y_pred):
     >>> y_true_imbalanced = [1, 2, 2, 2]
     >>> y_pred = [1, 2, 1, 2]
     >>> mean_absolute_error(y_true_balanced, y_pred)
-       0.5
+    0.5
     >>> mean_absolute_error(y_true_imbalanced, y_pred)
-       0.25
+    0.25
     >>> macro_averaged_mean_absolute_error(y_true_balanced, y_pred)
-       0.5
+    0.5
     >>> macro_averaged_mean_absolute_error(y_true_imbalanced, y_pred)
-       0.16666666666666666
+    0.16666666666666666
 
     """
     all_mae = []
-    y_true = np.array(y_true)
-    y_pred = np.array(y_pred)
+    y_true, y_pred = np.asarray(y_true), np.asarray(y_pred)
     for class_to_predict in np.unique(y_true):
-        index_class_to_predict = np.where(y_true == class_to_predict)[0]
+        index_class_to_predict = np.flatnonzero(y_true == class_to_predict)
         mae_class = mean_absolute_error(y_true[index_class_to_predict],
                                         y_pred[index_class_to_predict])
         all_mae.append(mae_class)