Merge pull request #2603 from ales-erjavec/fixes/randomization-seed-r…

…euse [FIX] preprocess.randomization: Do not use the same seed for X, Y, and meta
biolab · Sep 21, 2017 · 351b364 · 351b364
2 parents cdddfc3 + 8bdcb3e
commit 351b364
Show file tree

Hide file tree

Showing 2 changed files with 19 additions and 11 deletions.
diff --git a/Orange/preprocess/preprocess.py b/Orange/preprocess/preprocess.py
@@ -354,30 +354,33 @@ def __call__(self, data):
             Randomized data table.
         """
         new_data = data.copy()
+        rstate = np.random.RandomState(self.rand_seed)
+        # ensure the same seed is not used to shuffle X and Y at the same time
+        r1, r2, r3 = rstate.randint(0, 2 ** 32 - 1, size=3, dtype=np.int64)
         if self.rand_type & Randomize.RandomizeClasses:
-            new_data.Y = self.randomize(new_data.Y)
+            new_data.Y = self.randomize(new_data.Y, r1)
         if self.rand_type & Randomize.RandomizeAttributes:
-            new_data.X = self.randomize(new_data.X)
+            new_data.X = self.randomize(new_data.X, r2)
         if self.rand_type & Randomize.RandomizeMetas:
-            new_data.metas = self.randomize(new_data.metas)
+            new_data.metas = self.randomize(new_data.metas, r3)
         return new_data
 
-    def randomize(self, table):
-        np.random.seed(self.rand_seed)
+    def randomize(self, table, rand_state=None):
+        rstate = np.random.RandomState(rand_state)
         if sp.issparse(table):
             table = table.tocsc()
             rnd_indices = np.arange(table.shape[0], dtype=table.indices.dtype)
             for i in range(table.shape[1]):
                 col_indices = \
                     table.indices[table.indptr[i]: table.indptr[i + 1]]
                 new_indices = rnd_indices[:len(col_indices)]
-                np.random.shuffle(new_indices)
+                rstate.shuffle(new_indices)
                 col_indices[:] = new_indices
         elif len(table.shape) > 1:
             for i in range(table.shape[1]):
-                np.random.shuffle(table[:, i])
+                rstate.shuffle(table[:, i])
         else:
-            np.random.shuffle(table)
+            rstate.shuffle(table)
         return table
 
 

diff --git a/Orange/widgets/data/tests/test_owpreprocess.py b/Orange/widgets/data/tests/test_owpreprocess.py
@@ -25,11 +25,16 @@ def test_randomize(self):
         self.widget.set_model(model)
         self.send_signal(self.widget.Inputs.data, self.zoo)
         output = self.get_output(self.widget.Outputs.preprocessed_data)
-        np.random.seed(1)
-        np.random.shuffle(self.zoo.Y)
+        r = Randomize(Randomize.RandomizeClasses, rand_seed=1)
+        expected = r(self.zoo)
+
+        np.testing.assert_array_equal(expected.X, output.X)
+        np.testing.assert_array_equal(expected.Y, output.Y)
+        np.testing.assert_array_equal(expected.metas, output.metas)
+
         np.testing.assert_array_equal(self.zoo.X, output.X)
-        np.testing.assert_array_equal(self.zoo.Y, output.Y)
         np.testing.assert_array_equal(self.zoo.metas, output.metas)
+        self.assertFalse(np.array_equal(self.zoo.Y, output.Y))
 
     def test_normalize(self):
         data = Table("iris")