TestOWRandomForestRegression: Test upgrade

Author: VesnaT

Orange/widgets/regression/tests/test_owrandomforestregression.py

@@ -2,11 +2,61 @@
 # pylint: disable=missing-docstring
 from Orange.widgets.regression.owrandomforestregression import \
     OWRandomForestRegression
-from Orange.widgets.tests.base import WidgetTest, WidgetLearnerTestMixin
+from Orange.widgets.tests.base import (WidgetTest, WidgetLearnerTestMixin,
+                                       GuiToParam)
 
 
 class TestOWRandomForestRegression(WidgetTest, WidgetLearnerTestMixin):
     def setUp(self):
         self.widget = self.create_widget(OWRandomForestRegression,
                                          stored_settings={"auto_apply": False})
         self.init()
+        n_est_spin = self.widget.n_estimators_spin
+        max_f_spin = self.widget.max_features_spin[1]
+        rs_spin = self.widget.random_state_spin[1]
+        max_d_spin = self.widget.max_depth_spin[1]
+        min_s_spin = self.widget.min_samples_split_spin[1]
+        n_est_min_max = [n_est_spin.minimum() * 10, n_est_spin.minimum()]
+        min_s_min_max = [min_s_spin.minimum(), min_s_spin.maximum()]
+        self.gui_to_params = [
+            GuiToParam("n_estimators", n_est_spin, lambda x: x.value(),
+                       lambda i, x: x.setValue(i), n_est_min_max, n_est_min_max),
+            GuiToParam("max_features", max_f_spin, lambda x: "auto",
+                       lambda i, x: x.setValue(i), ["auto"], [0]),
+            GuiToParam("random_state", rs_spin, lambda x: None,
+                       lambda i, x: x.setValue(i), [None], [0]),
+            GuiToParam("max_depth", max_d_spin, lambda x: None,
+                       lambda i, x: x.setValue(i), [None], [0]),
+            GuiToParam("min_samples_split", min_s_spin, lambda x: x.value(),
+                       lambda i, x: x.setValue(i), min_s_min_max, min_s_min_max)]
+
+    def test_parameters_checked(self):
+        """Check learner and model for various values of all parameters
+        when all properties are checked
+        """
+        self.widget.max_features_spin[0].click()
+        self.widget.random_state_spin[0].click()
+        self.widget.max_depth_spin[0].click()
+        for j in range(1, 4):
+            el = self.gui_to_params[j]
+            el_min_max = [el.gui_el.minimum(), el.gui_el.maximum()]
+            self.gui_to_params[j] = GuiToParam(
+                el.name, el.gui_el, lambda x: x.value(),
+                lambda i, x: x.setValue(i), el_min_max, el_min_max)
+        self.test_parameters()
+        # FIXME: checkboxes are reset to default, since the widget settings were saved
+        self.widget.max_features_spin[0].setCheckState(False)
+        self.widget.random_state_spin[0].setCheckState(False)
+        self.widget.max_depth_spin[0].setCheckState(False)
+
+    def test_parameters_unchecked(self):
+        """Check learner and model for various values of all parameters
+        when properties are not checked
+        """
+        self.widget.min_samples_split_spin[0].click()
+        el = self.gui_to_params[4]
+        self.gui_to_params[4] = GuiToParam(el.name, el.gui_el, lambda x: 2,
+                                           lambda i, x: x.setValue(i), [2], [0])
+        self.test_parameters()
+        # FIXME: checkboxes are reset to default, since the widget settings were saved
+        self.widget.min_samples_split_spin[0].setCheckState(True)