first

aucpr.py

@@ -15,6 +15,7 @@
 
 import evaluate
 import datasets
+from sklearn.metrics import precision_recall_curve, auc
 
 
 # TODO: Add BibTeX citation
@@ -70,9 +71,13 @@ class AUCPR(evaluate.Metric):
             citation=_CITATION,
             inputs_description=_KWARGS_DESCRIPTION,
             # This defines the format of each prediction and reference
+#             features=datasets.Features({
+#                 'predictions': datasets.Value('int64'),
+#                 'references': datasets.Value('int64'),
+#             }),
             features=datasets.Features({
-                'predictions': datasets.Value('int64'),
-                'references': datasets.Value('int64'),
+                "prediction_scores": datasets.Value("float"),
+                "references": datasets.Value("int32"),
             }),
             # Homepage of the module for documentation
             homepage="http://module.homepage",
@@ -86,10 +91,11 @@ class AUCPR(evaluate.Metric):
         # TODO: Download external resources if needed
         pass
 
-    def _compute(self, predictions, references):
+    def _compute(self, references, prediction_scores):
         """Returns the scores"""
         # TODO: Compute the different scores of the module
-        accuracy = sum(i == j for i, j in zip(predictions, references)) / len(predictions)
+        precision, recall, _ = precision_recall_curve(references, prediction_scores)
+        aucpr = auc(recall, precision)
         return {
-            "accuracy": accuracy,
+            "aucpr": aucpr,
         }