def test_regression(self):
sf = StackedFitter([TreeLearner(), KNNLearner()])
cv = CrossValidation(k=3, random_state=0)
results = cv(self.housing[:50], [sf, TreeLearner(), KNNLearner()])
mse = MSE()(results)
self.assertLess(mse[0], mse[1])
self.assertLess(mse[0], mse[2])
def test_regression(self):
sf = StackedFitter([TreeLearner(), KNNLearner()])
results = CrossValidation(
self.housing[:50],
[sf, TreeLearner(), KNNLearner()], k=3)
mse = MSE(results)
self.assertLess(mse[0], mse[1])
self.assertLess(mse[0], mse[2])
def test_error_message_cleared_when_valid_learner_on_input(self):
# Disconnecting an invalid learner should use the default one and hide
# the error
self.send_signal("Learner", KNNLearner())
self.send_signal('Learner', None)
self.assertFalse(self.widget.Error.no_weight_support.is_shown(),
'Error message was not hidden on input disconnect')
# Connecting a valid learner should also reset the error message
self.send_signal("Learner", KNNLearner())
self.send_signal('Learner', RandomForestLearner())
self.assertFalse(
self.widget.Error.no_weight_support.is_shown(),
'Error message was not hidden when a valid learner appeared on '
'input')
def test_coverage(self):
# Due to async nature of these calls, these tests just cover
self.send_signal(self.widget.Inputs.data, self.data)
self.send_signal(self.widget.Inputs.learner, KNNLearner())
self.widget.class_attr = 'cls'
self.widget.handleNewSignals()
self.widget.map.fit_to_bounds()
self.widget.map.selected_area(90, 180, -90, -180)
self.widget.map.set_map_provider(
next(iter(self.widget.TILE_PROVIDERS.values())))
self.widget.map.set_clustering(True)
self.widget.map.set_clustering(False)
self.widget.map.set_jittering(5)
self.widget.map.set_marker_color('latitude')
self.widget.map.set_marker_label('latitude')
self.widget.map.set_marker_shape('foo')
self.widget.map.set_marker_size('latitude')
self.process_events(until=lambda start_time=time.clock(): time.clock()
- start_time > .2)
self.widget.map.set_marker_color(None)
self.widget.map.set_marker_label(None)
self.widget.map.set_marker_shape(None)
self.widget.map.set_marker_size(None)
self.widget.map.set_marker_size_coefficient(50)
self.widget.map.set_marker_opacity(20)
self.widget.map.recompute_heatmap(np.random.random((20, 2)))
args = [100, 100, -100, -100, 1000, 1000, 3, [-100, 100], [0, 0]]
self.widget.map.redraw_markers_overlay_image(*args, new_image=True)
# Force non-JS overlay redrawing
self.widget.map.N_POINTS_PER_ITER = 5
self.widget.map.redraw_markers_overlay_image(*args, new_image=True)
# pylint: disable=protected-access
assert (not np.isnan(self.widget.map._image_token)
and self.widget.map._image_token is not None)
self.process_events(until=lambda: self.widget.map._image_token is None)
self.widget.map.bridge.fit_to_bounds()
self.widget.map.bridge.selected_area(10, 20, 10, 20)
self.widget.map.bridge.recompute_heatmap(np.random.random((30, 2)))
self.widget.map.bridge.redraw_markers_overlay_image(
1, 2, 3, 4, 5, 6, 7, [1, 2], [3, 4])
self.widget.clear()
def test_input_learner_that_does_not_support_sample_weights(self):
self.send_signal("Learner", KNNLearner())
self.assertNotIsInstance(self.widget.base_estimator, KNNLearner)
self.assertIsNone(self.widget.base_estimator)
self.assertTrue(self.widget.Error.no_weight_support.is_shown())
def test_inputs(self):
self.send_signal(self.widget.Inputs.data, self.data)
self.send_signal(self.widget.Inputs.learner, KNNLearner())
self.widget.handleNewSignals()
self.assertEqual(self.widget.map.lat_attr, self.data.domain[0])
def test_classification(self):
sf = StackedFitter([TreeLearner(), KNNLearner()])
results = CrossValidation(self.iris, [sf], k=3)
ca = CA(results)
self.assertGreater(ca, 0.9)