distances: Fix tests for OWDistance

Author: janezd

Orange/widgets/unsupervised/tests/test_owdistances.py

@@ -5,7 +5,6 @@
 import numpy as np
 
 from Orange.data import Table
-from Orange.distance import MahalanobisDistance
 from Orange.widgets.unsupervised.owdistances import OWDistances, METRICS
 from Orange.widgets.tests.base import WidgetTest
 
@@ -24,9 +23,7 @@ def test_distance_combo(self):
         """Check distances when the metric changes"""
         self.assertEqual(self.widget.metrics_combo.count(), len(METRICS))
         self.send_signal(self.widget.Inputs.data, self.iris)
-        for i, metric in enumerate(METRICS):
-            if isinstance(metric, MahalanobisDistance):
-                metric = MahalanobisDistance(self.iris)
+        for i, (_, metric) in enumerate(METRICS):
             self.widget.metrics_combo.activated.emit(i)
             self.widget.metrics_combo.setCurrentIndex(i)
             self.send_signal(self.widget.Inputs.data, self.iris)
@@ -36,35 +33,14 @@ def test_distance_combo(self):
     def test_error_message(self):
         """Check if error message appears and then disappears when
         data is removed from input"""
+        self.widget.metric_idx = 2
         self.send_signal(self.widget.Inputs.data, self.iris)
         self.assertFalse(self.widget.Error.no_continuous_features.is_shown())
         self.send_signal(self.widget.Inputs.data, self.titanic)
         self.assertTrue(self.widget.Error.no_continuous_features.is_shown())
         self.send_signal(self.widget.Inputs.data, None)
         self.assertFalse(self.widget.Error.no_continuous_features.is_shown())
 
-    def test_mahalanobis_error(self):
-        mah_index = [i for i, d in enumerate(METRICS)
-                     if isinstance(d, MahalanobisDistance)][0]
-        self.widget.metric_idx = mah_index
-        self.widget.autocommit = True
-
-        invalid = self.iris[:]
-        invalid.X = np.vstack((invalid.X[0], invalid.X[0]))
-        invalid.Y = np.vstack((invalid.Y[0], invalid.Y[0]))
-        datasets = [self.iris, None, invalid]
-        bad = [False, False, True]
-        out = [True, False, False]
-
-        for data1, bad1, out1 in zip(datasets, bad, out):
-            for data2, bad2, out2 in zip(datasets, bad, out):
-                self.send_signal(self.widget.Inputs.data, data1)
-                self.assertEqual(self.widget.Error.mahalanobis_error.is_shown(), bad1)
-                self.assertEqual(self.get_output(self.widget.Outputs.distances) is not None, out1)
-                self.send_signal(self.widget.Inputs.data, data2)
-                self.assertEqual(self.widget.Error.mahalanobis_error.is_shown(), bad2)
-                self.assertEqual(self.get_output(self.widget.Outputs.distances) is not None, out2)
-
     def test_too_big_array(self):
         """
         Users sees an error message when calculating too large arrays and Orange