Merge pull request #1677 from VesnaT/annotated_data

[ENH] Visualize widgets: Output Annotated data and Fixups

Orange/widgets/data/owtable.py

@@ -34,6 +34,8 @@
 from Orange.widgets.settings import (Setting, ContextSetting,
                                      DomainContextHandler)
 from Orange.widgets.utils import datacaching
+from Orange.widgets.utils.annotated_data import (create_annotated_table,
+                                                 ANNOTATED_DATA_SIGNAL_NAME)
 from Orange.widgets.utils.itemmodels import TableModel
 
 
@@ -367,7 +369,7 @@ class OWDataTable(widget.OWWidget):
 
     inputs = [("Data", Table, "set_dataset", widget.Multiple)]
     outputs = [("Selected Data", Table, widget.Default),
-               ("Other Data", Table)]
+               (ANNOTATED_DATA_SIGNAL_NAME, Table)]
 
     show_distributions = Setting(False)
     dist_color_RGB = Setting((220, 220, 220, 255))
@@ -506,6 +508,7 @@ def update(f):
         self.selected_cols = []
         self.openContext(data)
         self.set_selection()
+        self.commit()
 
     def _setup_table_view(self, view, data):
         """Setup the `view` (QTableView) with `data` (Orange.data.Table)
@@ -785,7 +788,7 @@ def commit(self):
         """
         Commit/send the current selected row/column selection.
         """
-        selected_data = other_data = None
+        selected_data = table = rowsel = None
         view = self.tabs.currentWidget()
         if view and view.model() is not None:
             model = self._get_model(view)
@@ -795,7 +798,7 @@ def commit(self):
             # for SqlTables
             if isinstance(table, SqlTable):
                 self.send("Selected Data", selected_data)
-                self.send("Other Data", other_data)
+                self.send(ANNOTATED_DATA_SIGNAL_NAME, None)
                 return
 
             rowsel, colsel = self.get_selection(view)
@@ -841,19 +844,14 @@ def select_vars(role):
             # Avoid a copy if all/none rows are selected.
             if not rowsel:
                 selected_data = None
-                other_data = select(table, None, domain)
             elif len(rowsel) == len(table):
                 selected_data = select(table, None, domain)
-                other_data = None
             else:
                 selected_data = select(table, rowsel, domain)
-                selmask = numpy.ones((len(table),), dtype=bool)
-                selmask[rowsel] = False
-
-                other_data = select(table, numpy.flatnonzero(selmask), domain)
 
         self.send("Selected Data", selected_data)
-        self.send("Other Data", other_data)
+        self.send(ANNOTATED_DATA_SIGNAL_NAME,
+                  create_annotated_table(table, rowsel))
 
     def copy(self):
         """

Orange/widgets/data/tests/test_owtable.py

@@ -1,22 +1,37 @@
-from Orange.data import Table
+# Test methods with long descriptive names can omit docstrings
+# pylint: disable=missing-docstring
 from Orange.widgets.data.owtable import OWDataTable
-from Orange.widgets.tests.base import WidgetTest
+from Orange.widgets.tests.base import WidgetTest, WidgetOutputsTestMixin
 
 
-class TestOWDataTable(WidgetTest):
+class TestOWDataTable(WidgetTest, WidgetOutputsTestMixin):
+    @classmethod
+    def setUpClass(cls):
+        super().setUpClass()
+        WidgetOutputsTestMixin.init(cls)
+
+        cls.signal_name = "Data"
+        cls.signal_data = cls.data
+
     def setUp(self):
         self.widget = self.create_widget(OWDataTable)
-        self.iris = Table("iris")
 
     def test_input_data(self):
         """Check number of tabs with data on the input"""
-        self.send_signal("Data", self.iris, 1)
+        self.send_signal("Data", self.data, 1)
         self.assertEqual(self.widget.tabs.count(), 1)
-        self.send_signal("Data", self.iris, 2)
+        self.send_signal("Data", self.data, 2)
         self.assertEqual(self.widget.tabs.count(), 2)
         self.send_signal("Data", None, 1)
         self.assertEqual(self.widget.tabs.count(), 1)
 
     def test_data_model(self):
-        self.send_signal("Data", self.iris, 1)
-        self.assertEqual(self.widget.tabs.widget(0).model().rowCount(), len(self.iris))
+        self.send_signal("Data", self.data, 1)
+        self.assertEqual(self.widget.tabs.widget(0).model().rowCount(),
+                         len(self.data))
+
+    def _select_data(self):
+        self.widget.selected_cols = list(range(len(self.data.domain)))
+        self.widget.selected_rows = list(range(0, len(self.data.domain), 10))
+        self.widget.set_selection()
+        return self.widget.selected_rows

Orange/widgets/evaluate/tests/test_owconfusionmatrix.py

@@ -53,4 +53,4 @@ def _select_data(self):
         selected = [i for i, t in enumerate(zip(
             self.widget.results.actual, self.widget.results.predicted[0]))
                     if t in indices]
-        self.selected_indices = self.widget.results.row_indices[selected]
+        return self.widget.results.row_indices[selected]

Orange/widgets/tests/base.py

@@ -503,7 +503,6 @@ class WidgetOutputsTestMixin:
     def init(self):
         self.data = Table("iris")
         self.same_input_output_domain = True
-        self.selected_indices = []
 
     def test_outputs(self):
         self.send_signal(self.signal_name, self.signal_data)
@@ -523,7 +522,7 @@ def test_outputs(self):
         self.assertEqual(0, np.sum([i[feature_name] for i in annotated]))
 
         # select data instances
-        self._select_data()
+        selected_indices = self._select_data()
 
         # check selected data output
         selected = self.get_output("Selected Data")
@@ -532,7 +531,7 @@ def test_outputs(self):
         self.assertEqual(selected.domain == self.data.domain,
                          self.same_input_output_domain)
         np.testing.assert_array_equal(selected.X[:, :n_attr],
-                                      self.data.X[self.selected_indices])
+                                      self.data.X[selected_indices])
 
         # check annotated data output
         annotated = self.get_output(ANNOTATED_DATA_SIGNAL_NAME)
@@ -552,4 +551,4 @@ def _select_data(self):
     def _compare_selected_annotated_domains(self, selected, annotated):
         selected_vars = selected.domain.variables + selected.domain.metas
         annotated_vars = annotated.domain.variables + annotated.domain.metas
-        self.assertTrue(all((var in annotated_vars for var in selected_vars)))
+        self.assertLess(set(selected_vars), set(annotated_vars))

Orange/widgets/unsupervised/owhierarchicalclustering.py

@@ -1101,7 +1101,7 @@ def commit(self):
 
         if not selected_indices:
             self.send("Selected Data", None)
-            annotated_data = create_annotated_table(items, selected_indices) \
+            annotated_data = create_annotated_table(items, []) \
                 if self.selection_method == 0 and self.matrix.axis else None
             self.send(ANNOTATED_DATA_SIGNAL_NAME, annotated_data)
             return
@@ -1148,7 +1148,7 @@ def commit(self):
                 selected_data = data[mask]
                 if self.append_clusters:
                     def remove_other_value(vars_):
-                        vars_ = [var for var in vars_]
+                        vars_ = list(vars_)
                         clust_var = vars_[-1].copy()
                         clust_var.values.pop()
                         vars_[-1] = clust_var

Orange/widgets/unsupervised/tests/test_owdistancemap.py

@@ -20,6 +20,7 @@ def setUp(self):
 
     def _select_data(self):
         random.seed(42)
-        self.selected_indices = random.sample(range(0, len(self.data)), 20)
-        self.widget._selection = self.selected_indices
+        selected_indices = random.sample(range(0, len(self.data)), 20)
+        self.widget._selection = selected_indices
         self.widget.commit()
+        return selected_indices

Orange/widgets/unsupervised/tests/test_owhierarchicalclustering.py

@@ -25,16 +25,15 @@ def _select_data(self):
         items = self.widget.dendrogram._items
         cluster = items[sorted(list(items.keys()))[4]]
         self.widget.dendrogram.set_selected_items([cluster])
-        self.selected_indices = [14, 15, 32, 33]
+        return [14, 15, 32, 33]
 
     def _compare_selected_annotated_domains(self, selected, annotated):
-        self.assertTrue(all((var in annotated.domain.variables
-                             for var in selected.domain.variables)))
+        self.assertEqual(annotated.domain.variables,
+                         selected.domain.variables)
         self.assertNotIn("Other", selected.domain.metas[0].values)
         self.assertIn("Other", annotated.domain.metas[0].values)
-        self.assertTrue(
-            all((var in [var.name for var in annotated.domain.metas]
-                 for var in [var.name for var in selected.domain.metas])))
+        self.assertLess(set(var.name for var in selected.domain.metas),
+                        set(var.name for var in annotated.domain.metas))
 
     def test_selection_box_output(self):
         """Check output if Selection method changes"""

Orange/widgets/unsupervised/tests/test_owmds.py

@@ -25,4 +25,4 @@ def _select_data(self):
         points = random.sample(range(0, len(self.data)), 20)
         self.widget.select_indices(points)
         self.widget.commit()
-        self.selected_indices = sorted(points)
+        return sorted(points)

Orange/widgets/visualize/owlinearprojection.py

@@ -31,6 +31,9 @@
 from Orange.data.sql.table import SqlTable
 from Orange.widgets import widget, gui, settings
 from Orange.widgets.utils import itemmodels, colorpalette
+from Orange.widgets.utils.annotated_data import (
+    create_annotated_table, ANNOTATED_DATA_SIGNAL_NAME
+)
 from .owscatterplotgraph import LegendItem, legend_anchor_pos
 from Orange.widgets.utils import classdensity
 from Orange.canvas import report
@@ -238,7 +241,8 @@ class OWLinearProjection(widget.OWWidget):
               ("Data Subset", Table, "set_subset_data")]
 #              #TODO: Allow for axes to be supplied from an external source.
 #               ("Projection", numpy.ndarray, "set_axes"),]
-    outputs = [("Selected Data", Table)]
+    outputs = [("Selected Data", Table, widget.Default),
+               (ANNOTATED_DATA_SIGNAL_NAME, Table)]
 
     settingsHandler = settings.DomainContextHandler()
 
@@ -1064,12 +1068,15 @@ def select_indices(self, indices, modifiers=Qt.NoModifier):
 
     def commit(self):
         subset = None
+        indices = None
         if self.data is not None and self._selection_mask is not None:
             indices = numpy.flatnonzero(self._selection_mask)
             if len(indices) > 0:
                 subset = self.data[indices]
 
         self.send("Selected Data", subset)
+        self.send(ANNOTATED_DATA_SIGNAL_NAME,
+                  create_annotated_table(self.data, indices))
 
     def send_report(self):
         self.report_plot(name="", plot=self.viewbox.getViewBox())

Orange/widgets/visualize/owmosaic.py

@@ -13,10 +13,12 @@
 from Orange.preprocess import Discretize
 from Orange.preprocess.discretize import EqualFreq
 from Orange.statistics.distribution import get_distribution
-from Orange.widgets import gui
+from Orange.widgets import gui, widget
 from Orange.widgets.settings import (
     Setting, DomainContextHandler, ContextSetting)
 from Orange.widgets.utils import to_html, get_variable_values_sorted
+from Orange.widgets.utils.annotated_data import (create_annotated_table,
+                                                 ANNOTATED_DATA_SIGNAL_NAME)
 from Orange.widgets.visualize.utils import (
     CanvasText, CanvasRectangle, ViewWithPress)
 from Orange.widgets.widget import OWWidget, Default, Msg
@@ -30,7 +32,8 @@ class OWMosaicDisplay(OWWidget):
 
     inputs = [("Data", Table, "set_data", Default),
               ("Data Subset", Table, "set_subset_data")]
-    outputs = [("Selected Data", Table)]
+    outputs = [("Selected Data", Table, widget.Default),
+               (ANNOTATED_DATA_SIGNAL_NAME, Table)]
 
     settingsHandler = DomainContextHandler()
     use_boxes = Setting(True)
@@ -218,6 +221,8 @@ def select_area(self, index, ev):
     def send_selection(self):
         if not self.selection or self.data is None:
             self.send("Selected Data", None)
+            self.send(ANNOTATED_DATA_SIGNAL_NAME,
+                      create_annotated_table(self.data, []))
             return
         filters = []
         self.Warning.no_cont_selection_sql.clear()
@@ -235,12 +240,13 @@ def send_selection(self):
         else:
             filters = filters[0]
         selection = filters(self.discrete_data)
+        idset = set(selection.ids)
+        sel_idx = [i for i, id in enumerate(self.data.ids) if id in idset]
         if self.discrete_data is not self.data:
-            idset = set(selection.ids)
-            sel_idx = [i for i, id in enumerate(self.data.ids) if id in idset]
             selection = self.data[sel_idx]
         self.send("Selected Data", selection)
-
+        self.send(ANNOTATED_DATA_SIGNAL_NAME,
+                  create_annotated_table(self.data, sel_idx))
 
     def send_report(self):
         self.report_plot(self.canvas)

Orange/widgets/visualize/owpythagorastree.py

@@ -35,8 +35,12 @@
 )
 
 from Orange.data.table import Table
-from Orange.widgets import gui, settings
+from Orange.widgets import gui, settings, widget
 from Orange.widgets.utils import to_html
+from Orange.widgets.utils.annotated_data import (
+    create_annotated_table,
+    ANNOTATED_DATA_SIGNAL_NAME
+)
 from Orange.widgets.utils.colorpalette import ContinuousPaletteGenerator
 from Orange.widgets.visualize.pythagorastreeviewer import (
     PythagorasTreeViewer,
@@ -60,7 +64,8 @@ class OWPythagorasTree(OWWidget):
     priority = 1000
 
     inputs = [('Tree', TreeModel, 'set_tree')]
-    outputs = [('Selected Data', Table)]
+    outputs = [('Selected Data', Table, widget.Default),
+               (ANNOTATED_DATA_SIGNAL_NAME, Table)]
 
     # Enable the save as feature
     graph_name = 'scene'
@@ -228,6 +233,8 @@ def set_tree(self, model=None):
             # if hasattr(model, 'meta_depth_limit'):
             #     self.depth_limit = model.meta_depth_limit
             #     self.update_depth()
+        self.send(ANNOTATED_DATA_SIGNAL_NAME,
+                  create_annotated_table(self.instances, None))
 
     def clear(self):
         """Clear all relevant data from the widget."""
@@ -343,14 +350,16 @@ def commit(self):
         """Commit the selected data to output."""
         if self.instances is None:
             self.send('Selected Data', None)
+            self.send(ANNOTATED_DATA_SIGNAL_NAME, None)
             return
-        # this is taken almost directly from the owclassificationtreegraph.py
-        items = filter(lambda x: isinstance(x, SquareGraphicsItem),
-                       self.scene.selectedItems())
-
+        nodes = [i.tree_node.label for i in self.scene.selectedItems()
+                 if isinstance(i, SquareGraphicsItem)]
         data = self.tree_adapter.get_instances_in_nodes(
-            self.clf_dataset, [item.tree_node.label for item in items])
+            self.clf_dataset, nodes)
         self.send('Selected Data', data)
+        selected_indices = self.model.get_indices(nodes)
+        self.send(ANNOTATED_DATA_SIGNAL_NAME,
+                  create_annotated_table(self.instances, selected_indices))
 
     def send_report(self):
         """Send report."""