finished

sklearn_proxy.py

@@ -16,6 +16,7 @@
 import evaluate
 import datasets
 from sklearn.metrics import get_scorer
+from sklearn.base import BaseEstimator
 
 
 # TODO: Add BibTeX citation
@@ -61,6 +62,9 @@ BAD_WORDS_URL = "http://url/to/external/resource/bad_words.txt"
 @evaluate.utils.file_utils.add_start_docstrings(_DESCRIPTION, _KWARGS_DESCRIPTION)
 class SklearnProxy(evaluate.Metric):
     """TODO: Short description of my evaluation module."""
+    def __init__(self, **kwargs):
+        super().__init__(**kwargs)
+        self.dummy_estimator = PassThroughtEstimator()
 
     def _info(self):
         # TODO: Specifies the evaluate.EvaluationModuleInfo object
@@ -74,6 +78,7 @@ class SklearnProxy(evaluate.Metric):
             features=datasets.Features({
                 'predictions': datasets.Value('int64'),
                 'references': datasets.Value('int64'),
+
             }),
             # Homepage of the module for documentation
             homepage="http://module.homepage",
@@ -89,4 +94,14 @@ class SklearnProxy(evaluate.Metric):
 
     def _compute(self, predictions, references, metric_name="accuracy", **kwargs):
         scorer = get_scorer(metric_name)
-        return scorer(references, predictions, **kwargs)
+
+        return {metric_name: scorer(self.dummy_estimator, references, predictions, **kwargs)}
+
+    
+class PassThroughtEstimator(BaseEstimator):
+    def __init__(self):
+        pass
+    def fit(self, X, y):
+        return self
+    def predict(self, X):
+        return X

tests.py

@@ -1,17 +1,30 @@
-test_cases = [
+from sklearn_proxy import SklearnProxy
+import unittest
+
+
+accuracy_test_cases = [
     {
         "predictions": [0, 0],
         "references": [1, 1],
-        "result": {"metric_score": 0}
+        "result": {"accuracy": 0.0}
     },
     {
         "predictions": [1, 1],
         "references": [1, 1],
-        "result": {"metric_score": 1}
+        "result": {"accuracy": 1.0}
     },
     {
         "predictions": [1, 0],
         "references": [1, 1],
-        "result": {"metric_score": 0.5}
+        "result": {"accuracy": 0.5}
     }
-]
+]
+
+
+class TestGeneral(unittest.TestCase):
+
+    def test_accuracy(self):
+        metric = SklearnProxy()
+        for test_case in accuracy_test_cases:
+            result = metric.compute(predictions=test_case["predictions"],references=test_case["references"], metric_name="accuracy")
+            self.assertEqual(result, test_case["result"])