Transform: Replace 'Preprocess' input with 'Template Data' input

Author: VesnaT

Orange/widgets/data/owtransform.py

@@ -1,5 +1,5 @@
+import numpy as np
 from Orange.data import Table, Domain
-from Orange.preprocess.preprocess import Preprocess, Discretize
 from Orange.widgets import gui
 from Orange.widgets.settings import Setting
 from Orange.widgets.utils.sql import check_sql_input
@@ -18,29 +18,29 @@ class OWTransform(OWWidget):
 
     class Inputs:
         data = Input("Data", Table, default=True)
-        preprocessor = Input("Preprocessor", Preprocess)
+        template_data = Input("Template Data", Table)
 
     class Outputs:
         transformed_data = Output("Transformed Data", Table)
 
     class Error(OWWidget.Error):
-        pp_error = Msg("An error occurred while transforming data.\n{}")
+        error = Msg("An error occurred while transforming data.\n{}")
 
     resizing_enabled = False
     want_main_area = False
 
     def __init__(self):
         super().__init__()
         self.data = None
-        self.preprocessor = None
+        self.template_domain = None
         self.transformed_data = None
 
         info_box = gui.widgetBox(self.controlArea, "Info")
         self.input_label = gui.widgetLabel(info_box, "")
-        self.preprocessor_label = gui.widgetLabel(info_box, "")
+        self.template_label = gui.widgetLabel(info_box, "")
         self.output_label = gui.widgetLabel(info_box, "")
         self.set_input_label_text()
-        self.set_preprocessor_label_text()
+        self.set_template_label_text()
 
         box = gui.widgetBox(self.controlArea, "Output")
         gui.checkBox(box, self, "retain_all_data", "Retain all data",
@@ -54,13 +54,14 @@ def set_input_label_text(self):
                 len(self.data.domain.attributes))
         self.input_label.setText(text)
 
-    def set_preprocessor_label_text(self):
-        text = "No preprocessor on input."
+    def set_template_label_text(self):
+        text = "No template data on input."
         if self.transformed_data is not None:
-            text = "Preprocessor {} applied.".format(self.preprocessor)
-        elif self.preprocessor is not None:
-            text = "Preprocessor {} on input.".format(self.preprocessor)
-        self.preprocessor_label.setText(text)
+            text = "Template domain applied."
+        elif self.template_domain is not None:
+            text = "Template data includes {:,} features.".format(
+                len(self.template_domain.attributes))
+        self.template_label.setText(text)
 
     def set_output_label_text(self):
         text = ""
@@ -75,49 +76,55 @@ def set_data(self, data):
         self.data = data
         self.set_input_label_text()
 
-    @Inputs.preprocessor
-    def set_preprocessor(self, preprocessor):
-        self.preprocessor = preprocessor
+    @Inputs.template_data
+    @check_sql_input
+    def set_template_data(self, data):
+        self.template_domain = data and data.domain
 
     def handleNewSignals(self):
         self.apply()
 
     def apply(self):
         self.clear_messages()
         self.transformed_data = None
-        if self.data is not None and self.preprocessor is not None:
+        if self.data is not None and self.template_domain is not None:
             try:
-                self.transformed_data = self.preprocessor(self.data)
-            except Exception as ex:   # pylint: disable=broad-except
-                self.Error.pp_error(ex)
+                self.transformed_data = self.data.transform(
+                    self.template_domain)
+            except Exception as ex:  # pylint: disable=broad-except
+                self.Error.error(ex)
 
         data = self.merged_data() if self.retain_all_data \
             else self.transformed_data
         self.Outputs.transformed_data.send(data)
-
-        self.set_preprocessor_label_text()
+        self.set_template_label_text()
         self.set_output_label_text()
 
     def merged_data(self):
+        if not self.transformed_data:
+            return None
         domain = self.data.domain
-        metas = domain.metas + self.transformed_data.domain.attributes
+        t_domain = self.transformed_data.domain
+        metas = domain.metas + t_domain.attributes + t_domain.metas
         domain = Domain(domain.attributes, domain.class_vars, metas)
         data = self.data.transform(domain)
-        n = self.transformed_data.X.shape[1]
-        data.metas[:, -n:] = self.transformed_data.X
+        metas = np.hstack((self.transformed_data.X,
+                           self.transformed_data.metas))
+        data.metas[:, -metas.shape[1]:] = metas
         return data
 
     def send_report(self):
-        if self.preprocessor is not None:
-            self.report_items("Settings",
-                              (("Preprocessor", self.preprocessor),))
         if self.data is not None:
             self.report_data("Data", self.data)
+        if self.template_domain is not None:
+            self.report_domain("Template data", self.template_domain)
         if self.transformed_data is not None:
             self.report_data("Transformed data", self.transformed_data)
 
 
 if __name__ == "__main__":  # pragma: no cover
+    from Orange.preprocess import Discretize
+
+    table = Table("iris")
     WidgetPreview(OWTransform).run(
-        set_data=Table("iris"),
-        set_preprocessor=Discretize())
+        set_data=table, set_template_data=Discretize()(table))

Orange/widgets/data/tests/test_owtransform.py

@@ -1,8 +1,11 @@
 # Test methods with long descriptive names can omit docstrings
 # pylint: disable=missing-docstring
+from unittest.mock import Mock
+
+from numpy import testing as npt
+
 from Orange.data import Table
 from Orange.preprocess import Discretize, Continuize
-from Orange.preprocess.preprocess import Preprocess
 from Orange.widgets.data.owtransform import OWTransform
 from Orange.widgets.tests.base import WidgetTest
 from Orange.widgets.unsupervised.owpca import OWPCA
@@ -12,39 +15,39 @@ class TestOWTransform(WidgetTest):
     def setUp(self):
         self.widget = self.create_widget(OWTransform)
         self.data = Table("iris")
-        self.preprocessor = Discretize()
+        self.disc_data = Discretize()(self.data)
 
     def test_output(self):
-        # send data and preprocessor
-        self.send_signal(self.widget.Inputs.data, self.data)
-        self.send_signal(self.widget.Inputs.preprocessor, self.preprocessor)
+        # send data and template data
+        self.send_signal(self.widget.Inputs.data, self.data[::15])
+        self.send_signal(self.widget.Inputs.template_data, self.disc_data)
         output = self.get_output(self.widget.Outputs.transformed_data)
-        self.assertIsInstance(output, Table)
-        self.assertEqual("Input data with 150 instances and 4 features.",
+        self.assertTableEqual(output, self.disc_data[::15])
+        self.assertEqual("Input data with 10 instances and 4 features.",
                          self.widget.input_label.text())
-        self.assertEqual("Preprocessor Discretize() applied.",
-                         self.widget.preprocessor_label.text())
+        self.assertEqual("Template domain applied.",
+                         self.widget.template_label.text())
         self.assertEqual("Output data includes 4 features.",
                          self.widget.output_label.text())
 
-        # remove preprocessor
-        self.send_signal(self.widget.Inputs.preprocessor, None)
+        # remove template data
+        self.send_signal(self.widget.Inputs.template_data, None)
         output = self.get_output(self.widget.Outputs.transformed_data)
         self.assertIsNone(output)
-        self.assertEqual("Input data with 150 instances and 4 features.",
+        self.assertEqual("Input data with 10 instances and 4 features.",
                          self.widget.input_label.text())
-        self.assertEqual("No preprocessor on input.",
-                         self.widget.preprocessor_label.text())
+        self.assertEqual("No template data on input.",
+                         self.widget.template_label.text())
         self.assertEqual("", self.widget.output_label.text())
 
-        # send preprocessor
-        self.send_signal(self.widget.Inputs.preprocessor, self.preprocessor)
+        # send template data
+        self.send_signal(self.widget.Inputs.template_data, self.disc_data)
         output = self.get_output(self.widget.Outputs.transformed_data)
-        self.assertIsInstance(output, Table)
-        self.assertEqual("Input data with 150 instances and 4 features.",
+        self.assertTableEqual(output, self.disc_data[::15])
+        self.assertEqual("Input data with 10 instances and 4 features.",
                          self.widget.input_label.text())
-        self.assertEqual("Preprocessor Discretize() applied.",
-                         self.widget.preprocessor_label.text())
+        self.assertEqual("Template domain applied.",
+                         self.widget.template_label.text())
         self.assertEqual("Output data includes 4 features.",
                          self.widget.output_label.text())
 
@@ -53,48 +56,55 @@ def test_output(self):
         output = self.get_output(self.widget.Outputs.transformed_data)
         self.assertIsNone(output)
         self.assertEqual("No data on input.", self.widget.input_label.text())
-        self.assertEqual("Preprocessor Discretize() on input.",
-                         self.widget.preprocessor_label.text())
+        self.assertEqual("Template data includes 4 features.",
+                         self.widget.template_label.text())
         self.assertEqual("", self.widget.output_label.text())
 
-        # remove preprocessor
-        self.send_signal(self.widget.Inputs.preprocessor, None)
+        # remove template data
+        self.send_signal(self.widget.Inputs.template_data, None)
         self.assertEqual("No data on input.", self.widget.input_label.text())
-        self.assertEqual("No preprocessor on input.",
-                         self.widget.preprocessor_label.text())
+        self.assertEqual("No template data on input.",
+                         self.widget.template_label.text())
         self.assertEqual("", self.widget.output_label.text())
 
-    def test_input_pca_preprocessor(self):
+    def assertTableEqual(self, table1, table2):
+        self.assertIs(table1.domain, table2.domain)
+        npt.assert_array_equal(table1.X, table2.X)
+        npt.assert_array_equal(table1.Y, table2.Y)
+        npt.assert_array_equal(table1.metas, table2.metas)
+
+    def test_input_pca_output(self):
         owpca = self.create_widget(OWPCA)
         self.send_signal(owpca.Inputs.data, self.data, widget=owpca)
         owpca.components_spin.setValue(2)
-        pp = self.get_output(owpca.Outputs.preprocessor, widget=owpca)
-        self.assertIsNotNone(pp, Preprocess)
+        pca_out = self.get_output(owpca.Outputs.transformed_data, widget=owpca)
 
-        self.send_signal(self.widget.Inputs.data, self.data)
-        self.send_signal(self.widget.Inputs.preprocessor, pp)
+        self.send_signal(self.widget.Inputs.data, self.data[::10])
+        self.send_signal(self.widget.Inputs.template_data, pca_out)
         output = self.get_output(self.widget.Outputs.transformed_data)
-        self.assertIsInstance(output, Table)
-        self.assertEqual(output.X.shape, (len(self.data), 2))
+        npt.assert_array_equal(pca_out.X[::10], output.X)
 
     def test_retain_all_data(self):
         data = Table("zoo")
+        cont_data = Continuize()(data)
         self.send_signal(self.widget.Inputs.data, data)
-        self.send_signal(self.widget.Inputs.preprocessor, Continuize())
+        self.send_signal(self.widget.Inputs.template_data, cont_data)
         self.widget.controls.retain_all_data.click()
         output = self.get_output(self.widget.Outputs.transformed_data)
         self.assertIsInstance(output, Table)
         self.assertEqual(output.X.shape, (len(data), 16))
-        self.assertEqual(output.metas.shape, (len(data), 37))
+        self.assertEqual(output.metas.shape, (len(data), 38))
 
     def test_error_transforming(self):
-        self.send_signal(self.widget.Inputs.data, self.data)
-        self.send_signal(self.widget.Inputs.preprocessor, Preprocess())
-        self.assertTrue(self.widget.Error.pp_error.is_shown())
+        data = self.data[::10]
+        data.transform = Mock(side_effect=Exception())
+        self.send_signal(self.widget.Inputs.data, data)
+        self.send_signal(self.widget.Inputs.template_data, self.disc_data)
+        self.assertTrue(self.widget.Error.error.is_shown())
         output = self.get_output(self.widget.Outputs.transformed_data)
         self.assertIsNone(output)
         self.send_signal(self.widget.Inputs.data, None)
-        self.assertFalse(self.widget.Error.pp_error.is_shown())
+        self.assertFalse(self.widget.Error.error.is_shown())
 
     def test_send_report(self):
         self.send_signal(self.widget.Inputs.data, self.data)