Unittests: Refactoring

Orange/widgets/tests/base.py

@@ -1,10 +1,13 @@
 import unittest
 
+import numpy as np
+
 from PyQt4.QtGui import (QApplication, QComboBox, QSpinBox, QDoubleSpinBox,
                          QSlider)
 import sip
 
 from Orange.data import Table
+from Orange.misc.flagged_data import FLAGGED_SIGNAL_NAME, FLAGGED_FEATURE_NAME
 from Orange.preprocess import RemoveNaNColumns, Randomize
 from Orange.preprocess.preprocess import PreprocessorList
 from Orange.classification.base_classification import (LearnerClassification,
@@ -453,3 +456,70 @@ def test_parameters(self):
                     self.assertFalse(self.widget.Error.active)
                 else:
                     self.assertTrue(self.widget.Error.active)
+
+
+class WidgetOutputsTestMixin:
+    """Class for widget's outputs testing.
+
+    Contains init method to set up testing parameters and a test method, which
+    checks Selected Data and Flagged Data outputs.
+
+    Since widgets have different ways of selecting data instances, _select_data
+    method should be implemented when subclassed.
+
+    If output's expected domain differs from input's domain, parameter
+    same_input_output_domain should be set to False.
+
+    If Selected Data and Flagged Data domains differ, override method
+    _compare_selected_flagged_domains.
+    """
+
+    def init(self):
+        self.data = Table("iris")
+        self.same_input_output_domain = True
+
+    def test_outputs(self):
+        self.send_signal(self.signal_name, self.signal_data)
+
+        # only needed in TestOWMDS
+        if type(self).__name__ == "TestOWMDS":
+            from PyQt4.QtCore import QEvent
+            self.widget.customEvent(QEvent(QEvent.User))
+            self.widget.commit()
+
+        # check selected data output
+        self.assertIsNone(self.get_output("Selected Data"))
+
+        # check flagged data output
+        flagged = self.get_output(FLAGGED_SIGNAL_NAME)
+        self.assertEqual(0, np.sum([i[FLAGGED_FEATURE_NAME] for i in flagged]))
+
+        # select data instances
+        self._select_data()
+
+        # check selected data output
+        selected = self.get_output("Selected Data")
+        self.assertGreater(len(selected), 0)
+        if self.same_input_output_domain:
+            self.assertEqual(selected.domain, self.data.domain)
+
+        # check flagged data output
+        flagged = self.get_output(FLAGGED_SIGNAL_NAME)
+        self.assertEqual(len(selected),
+                         np.sum([i[FLAGGED_FEATURE_NAME] for i in flagged]))
+
+        # compare selected and flagged data domains
+        self._compare_selected_flagged_domains(selected, flagged)
+
+        # check output when data is removed
+        self.send_signal(self.signal_name, None)
+        self.assertIsNone(self.get_output("Selected Data"))
+        self.assertIsNone(self.get_output(FLAGGED_SIGNAL_NAME))
+
+    def _select_data(self):
+        raise NotImplementedError("Subclasses should implement select_data")
+
+    def _compare_selected_flagged_domains(self, selected, flagged):
+        selected_vars = selected.domain.variables + selected.domain.metas
+        flagged_vars = flagged.domain.variables + flagged.domain.metas
+        self.assertTrue(all((var in flagged_vars for var in selected_vars)))

Orange/widgets/unsupervised/tests/test_owdistancemap.py

@@ -1,50 +1,24 @@
 # Test methods with long descriptive names can omit docstrings
 # pylint: disable=missing-docstring
 import random
-import numpy as np
-from Orange.data import Table
-from Orange.misc.flagged_data import FLAGGED_SIGNAL_NAME, FLAGGED_FEATURE_NAME
 from Orange.distance import Euclidean
 from Orange.widgets.unsupervised.owdistancemap import OWDistanceMap
-from Orange.widgets.tests.base import WidgetTest
+from Orange.widgets.tests.base import WidgetTest, WidgetOutputsTestMixin
 
 
-class TestOWDistanceMap(WidgetTest):
-    def setUp(self):
-        self.widget = self.create_widget(OWDistanceMap)
-        self.iris = Table("iris")
-        self.iris_distances = Euclidean(self.iris)
-
-    def test_outputs(self):
-        self.send_signal("Distances", self.iris_distances)
+class TestOWDistanceMap(WidgetTest, WidgetOutputsTestMixin):
+    @classmethod
+    def setUpClass(cls):
+        super().setUpClass()
+        WidgetOutputsTestMixin.init(cls)
 
-        # check selected data output
-        self.assertIsNone(self.get_output("Selected Data"))
+        cls.signal_name = "Distances"
+        cls.signal_data = Euclidean(cls.data)
 
-        # check flagged data output
-        flagged = self.get_output(FLAGGED_SIGNAL_NAME)
-        self.assertEqual(0, np.sum([i[FLAGGED_FEATURE_NAME] for i in flagged]))
+    def setUp(self):
+        self.widget = self.create_widget(OWDistanceMap)
 
-        # select data points
-        points = random.sample(range(0, len(self.iris)), 20)
+    def _select_data(self):
+        points = random.sample(range(0, len(self.data)), 20)
         self.widget._selection = points
         self.widget.commit()
-
-        # check selected data output
-        selected = self.get_output("Selected Data")
-        self.assertEqual(len(selected), len(points))
-
-        # check flagged data output
-        flagged = self.get_output(FLAGGED_SIGNAL_NAME)
-        self.assertEqual(len(selected),
-                         np.sum([i[FLAGGED_FEATURE_NAME] for i in flagged]))
-
-        # compare selected and flagged data domains
-        selected_vars = selected.domain.variables + selected.domain.metas
-        flagged_vars = flagged.domain.variables + flagged.domain.metas
-        self.assertTrue(all((var in flagged_vars for var in selected_vars)))
-
-        # check output when data is removed
-        self.send_signal("Distances", None)
-        self.assertIsNone(self.get_output("Selected Data"))
-        self.assertIsNone(self.get_output(FLAGGED_SIGNAL_NAME))

Orange/widgets/unsupervised/tests/test_owhierarchicalclustering.py

@@ -1,45 +1,32 @@
 # Test methods with long descriptive names can omit docstrings
 # pylint: disable=missing-docstring
-import numpy as np
-from Orange.data import Table
-from Orange.misc.flagged_data import FLAGGED_SIGNAL_NAME, FLAGGED_FEATURE_NAME
+from Orange.misc.flagged_data import FLAGGED_SIGNAL_NAME
 from Orange.distance import Euclidean
 from Orange.widgets.unsupervised.owhierarchicalclustering import \
     OWHierarchicalClustering
-from Orange.widgets.tests.base import WidgetTest
+from Orange.widgets.tests.base import WidgetTest, WidgetOutputsTestMixin
 
 
-class TestOWHierarchicalClustering(WidgetTest):
-    def setUp(self):
-        self.widget = self.create_widget(OWHierarchicalClustering)
-        self.iris = Table("iris")
-        self.iris_distances = Euclidean(self.iris)
-
-    def test_outputs(self):
-        self.send_signal("Distances", self.iris_distances)
+class TestOWHierarchicalClustering(WidgetTest, WidgetOutputsTestMixin):
+    @classmethod
+    def setUpClass(cls):
+        super().setUpClass()
+        WidgetOutputsTestMixin.init(cls)
 
-        # check selected data output
-        self.assertIsNone(self.get_output("Selected Data"))
+        cls.distances = Euclidean(cls.data)
+        cls.signal_name = "Distances"
+        cls.signal_data = cls.distances
+        cls.same_input_output_domain = False
 
-        # check flagged data output
-        flagged = self.get_output(FLAGGED_SIGNAL_NAME)
-        self.assertEqual(0, np.sum([i[FLAGGED_FEATURE_NAME] for i in flagged]))
+    def setUp(self):
+        self.widget = self.create_widget(OWHierarchicalClustering)
 
-        # select a cluster
+    def _select_data(self):
         items = self.widget.dendrogram._items
         cluster = items[next(iter(items))]
         self.widget.dendrogram.set_selected_items([cluster])
 
-        # check selected data output
-        selected = self.get_output("Selected Data")
-        self.assertGreater(len(selected), 0)
-
-        # check flagged data output
-        flagged = self.get_output(FLAGGED_SIGNAL_NAME)
-        self.assertEqual(len(selected),
-                         np.sum([i[FLAGGED_FEATURE_NAME] for i in flagged]))
-
-        # compare selected and flagged data domains
+    def _compare_selected_flagged_domains(self, selected, flagged):
         self.assertTrue(all((var in flagged.domain.variables
                              for var in selected.domain.variables)))
         self.assertNotIn("Other", selected.domain.metas[0].values)
@@ -48,14 +35,9 @@ def test_outputs(self):
             all((var in [var.name for var in flagged.domain.metas]
                  for var in [var.name for var in selected.domain.metas])))
 
-        # check output when data is removed
-        self.send_signal("Distances", None)
-        self.assertIsNone(self.get_output("Selected Data"))
-        self.assertIsNone(self.get_output(FLAGGED_SIGNAL_NAME))
-
     def test_selection_box_output(self):
         """Check output if Selection method changes"""
-        self.send_signal("Distances", self.iris_distances)
+        self.send_signal("Distances", self.distances)
         self.assertIsNone(self.get_output("Selected Data"))
         self.assertIsNotNone(self.get_output(FLAGGED_SIGNAL_NAME))
 

Orange/widgets/unsupervised/tests/test_owmds.py

@@ -1,53 +1,25 @@
 # Test methods with long descriptive names can omit docstrings
 # pylint: disable=missing-docstring
 import random
-import numpy as np
-from PyQt4.QtCore import QEvent
-from Orange.data import Table
-from Orange.misc.flagged_data import FLAGGED_SIGNAL_NAME, FLAGGED_FEATURE_NAME
 from Orange.distance import Euclidean
 from Orange.widgets.unsupervised.owmds import OWMDS
-from Orange.widgets.tests.base import WidgetTest
+from Orange.widgets.tests.base import WidgetTest, WidgetOutputsTestMixin
 
 
-class TestOWMDS(WidgetTest):
-    def setUp(self):
-        self.widget = self.create_widget(OWMDS)
-        self.iris = Table("iris")
-        self.iris_distances = Euclidean(self.iris)
+class TestOWMDS(WidgetTest, WidgetOutputsTestMixin):
+    @classmethod
+    def setUpClass(cls):
+        super().setUpClass()
+        WidgetOutputsTestMixin.init(cls)
 
-    def test_outputs(self):
-        self.send_signal("Distances", self.iris_distances)
-        self.widget.customEvent(QEvent(QEvent.User))
-        self.widget.commit()
+        cls.signal_name = "Distances"
+        cls.signal_data = Euclidean(cls.data)
+        cls.same_input_output_domain = False
 
-        # check selected data output
-        self.assertIsNone(self.get_output("Selected Data"))
-
-        # check flagged data output
-        flagged = self.get_output(FLAGGED_SIGNAL_NAME)
-        self.assertEqual(0, np.sum([i[FLAGGED_FEATURE_NAME] for i in flagged]))
+    def setUp(self):
+        self.widget = self.create_widget(OWMDS)
 
-        # select data points
-        points = random.sample(range(0, len(self.iris)), 20)
+    def _select_data(self):
+        points = random.sample(range(0, len(self.data)), 20)
         self.widget.select_indices(points)
         self.widget.commit()
-
-        # check selected data output
-        selected = self.get_output("Selected Data")
-        self.assertEqual(len(selected), len(points))
-
-        # check flagged data output
-        flagged = self.get_output(FLAGGED_SIGNAL_NAME)
-        self.assertEqual(len(selected),
-                         np.sum([i[FLAGGED_FEATURE_NAME] for i in flagged]))
-
-        # compare selected and flagged data domains
-        selected_vars = selected.domain.variables + selected.domain.metas
-        flagged_vars = flagged.domain.variables + flagged.domain.metas
-        self.assertTrue(all((var in flagged_vars for var in selected_vars)))
-
-        # check output when data is removed
-        self.send_signal("Distances", None)
-        self.assertIsNone(self.get_output("Selected Data"))
-        self.assertIsNone(self.get_output(FLAGGED_SIGNAL_NAME))