Corrects the forgotten bits of PR #267 (#269)

* maj

* maj

* corrected PR 267

* trailing whitespace

* test calibrate_threshold, test predict

* maj

* Checks estimator is fitted before set threshold

* correct failed tests with MockBadClassifier

* remove checks

* forgot one

* missed one check_is_fitted

* sklearn changed the assumptions behind check_is_fitted

Author: RobinVogel

metric_learn/base_metric.py

@@ -93,6 +93,8 @@ def _prepare_inputs(self, X, y=None, type_of_inputs='classic',
       The checked input labels array.
     """
     self._check_preprocessor()
+
+    check_is_fitted(self, ['preprocessor_'])
     return check_input(X, y,
                        type_of_inputs=type_of_inputs,
                        preprocessor=self.preprocessor_,
@@ -215,6 +217,7 @@ def score_pairs(self, pairs):
     :ref:`mahalanobis_distances` : The section of the project documentation
       that describes Mahalanobis Distances.
     """
+    check_is_fitted(self, ['preprocessor_'])
     pairs = check_input(pairs, type_of_inputs='tuples',
                         preprocessor=self.preprocessor_,
                         estimator=self, tuple_size=2)
@@ -336,8 +339,10 @@ def predict(self, pairs):
     y_predicted : `numpy.ndarray` of floats, shape=(n_constraints,)
       The predicted learned metric value between samples in every pair.
     """
+    check_is_fitted(self, 'preprocessor_')
+
     if "threshold_" not in vars(self):
-      msg = ("A threshold for this estimator has not been set,"
+      msg = ("A threshold for this estimator has not been set, "
              "call its set_threshold or calibrate_threshold method.")
       raise AttributeError(msg)
     return 2 * (- self.decision_function(pairs) <= self.threshold_) - 1
@@ -414,6 +419,8 @@ def set_threshold(self, threshold):
     self : `_PairsClassifier`
       The pairs classifier with the new threshold set.
     """
+    check_is_fitted(self, 'preprocessor_')
+
     self.threshold_ = threshold
     return self
 
@@ -476,6 +483,7 @@ def calibrate_threshold(self, pairs_valid, y_valid, strategy='accuracy',
     --------
     sklearn.calibration : scikit-learn's module for calibrating classifiers
     """
+    check_is_fitted(self, 'preprocessor_')
 
     self._validate_calibration_params(strategy, min_rate, beta)
 

test/test_pairs_classifiers.py

@@ -66,14 +66,31 @@ def test_predict_monotonous(estimator, build_dataset,
                          ids=ids_pairs_learners)
 def test_raise_not_fitted_error_if_not_fitted(estimator, build_dataset,
                                               with_preprocessor):
-  """Test that a NotFittedError is raised if someone tries to predict and
-  the metric learner has not been fitted."""
+  """Test that a NotFittedError is raised if someone tries to use
+  score_pairs, decision_function, get_metric, transform or
+  get_mahalanobis_matrix on input data and the metric learner
+  has not been fitted."""
   input_data, labels, preprocessor, _ = build_dataset(with_preprocessor)
   estimator = clone(estimator)
   estimator.set_params(preprocessor=preprocessor)
   set_random_state(estimator)
+  with pytest.raises(NotFittedError):
+    estimator.score_pairs(input_data)
   with pytest.raises(NotFittedError):
     estimator.decision_function(input_data)
+  with pytest.raises(NotFittedError):
+    estimator.get_metric()
+  with pytest.raises(NotFittedError):
+    estimator.transform(input_data)
+  with pytest.raises(NotFittedError):
+    estimator.get_mahalanobis_matrix()
+  with pytest.raises(NotFittedError):
+    estimator.calibrate_threshold(input_data, labels)
+
+  with pytest.raises(NotFittedError):
+    estimator.set_threshold(0.5)
+  with pytest.raises(NotFittedError):
+    estimator.predict(input_data)
 
 
 @pytest.mark.parametrize('calibration_params',
@@ -138,15 +155,16 @@ def fit(self, pairs, y):
 
 
 def test_unset_threshold():
-  # test that set_threshold indeed sets the threshold
+  """Tests that the "threshold is unset" error is raised when using predict
+  (performs binary classification on pairs) with an unset threshold."""
   identity_pairs_classifier = IdentityPairsClassifier()
   pairs = np.array([[[0.], [1.]], [[1.], [3.]], [[2.], [5.]], [[3.], [7.]]])
   y = np.array([1, 1, -1, -1])
   identity_pairs_classifier.fit(pairs, y)
   with pytest.raises(AttributeError) as e:
     identity_pairs_classifier.predict(pairs)
 
-  expected_msg = ("A threshold for this estimator has not been set,"
+  expected_msg = ("A threshold for this estimator has not been set, "
                   "call its set_threshold or calibrate_threshold method.")
 
   assert str(e.value) == expected_msg
@@ -362,6 +380,7 @@ class MockBadPairsClassifier(MahalanobisMixin, _PairsClassifierMixin):
     """
 
     def fit(self, pairs, y, calibration_params=None):
+      self.preprocessor_ = 'not used'
       self.components_ = 'not used'
       self.calibrate_threshold(pairs, y, **(calibration_params if
                                             calibration_params is not None else

test/test_utils.py

@@ -749,6 +749,8 @@ def test_array_like_indexer_array_like_valid_classic(input_data, indices):
   """Checks that any array-like is valid in the 'preprocessor' argument,
   and in the indices, for a classic input"""
   class MockMetricLearner(MahalanobisMixin):
+    def fit(self):
+      pass
     pass
 
   mock_algo = MockMetricLearner(preprocessor=input_data)
@@ -763,6 +765,8 @@ def test_array_like_indexer_array_like_valid_tuples(input_data, indices):
   """Checks that any array-like is valid in the 'preprocessor' argument,
   and in the indices, for a classic input"""
   class MockMetricLearner(MahalanobisMixin):
+    def fit(self):
+      pass
     pass
 
   mock_algo = MockMetricLearner(preprocessor=input_data)