Merge pull request #6784 from VesnaT/pls_missing_target

PLS: Handle missing target values
biolab · May 10, 2024 · 4f4539a · 4f4539a
2 parents d8818b5 + 538e16d
commit 4f4539a
Show file tree

Hide file tree

Showing 3 changed files with 58 additions and 1 deletion.
diff --git a/Orange/regression/pls.py b/Orange/regression/pls.py
@@ -36,7 +36,16 @@ def _transform_with_numpy_output(self, X, Y):
         u = y_center @ pls.y_rotations_
         """
         pls = self.pls_model.skl_model
-        t, u = pls.transform(X, Y)
+        mask = np.isnan(Y).any(axis=1)
+        n_comp = pls.n_components
+        t = np.full((len(X), n_comp), np.nan, dtype=float)
+        u = np.full((len(X), n_comp), np.nan, dtype=float)
+        if (~mask).sum() > 0:
+            t_, u_ = pls.transform(X[~mask], Y[~mask])
+            t[~mask] = t_
+            u[~mask] = u_
+        if mask.sum() > 0:
+            t[mask] = pls.transform(X[mask])
         return np.hstack((t, u))
 
     def __call__(self, data):

diff --git a/Orange/regression/tests/test_pls.py b/Orange/regression/tests/test_pls.py
@@ -154,6 +154,38 @@ def test_hash(self):
         m = PLSRegressionLearner()(table(10, 5, 2))
         self.assertNotEqual(hash(transformer), hash(_PLSCommonTransform(m)))
 
+    def test_missing_target(self):
+        data = table(10, 5, 1)
+        with data.unlocked(data.Y):
+            data.Y[::3] = np.nan
+        pls = PLSRegressionLearner()(data)
+        proj = pls.project(data)
+        self.assertFalse(np.isnan(proj.X).any())
+        self.assertFalse(np.isnan(proj.metas[1::3]).any())
+        self.assertFalse(np.isnan(proj.metas[2::3]).any())
+        self.assertTrue(np.isnan(proj.metas[::3]).all())
+
+    def test_missing_target_multitarget(self):
+        data = table(10, 5, 3)
+        with data.unlocked(data.Y):
+            data.Y[0] = np.nan
+            data.Y[1, 1] = np.nan
+
+        pls = PLSRegressionLearner()(data)
+        proj = pls.project(data)
+        self.assertFalse(np.isnan(proj.X).any())
+        self.assertFalse(np.isnan(proj.metas[2:]).any())
+        self.assertTrue(np.isnan(proj.metas[:2]).all())
+
+    def test_apply_domain_classless_data(self):
+        data = Table("housing")
+        pls = PLSRegressionLearner()(data)
+        classless_data = data.transform(Domain(data.domain.attributes))[:5]
+
+        proj = pls.project(classless_data)
+        self.assertFalse(np.isnan(proj.X).any())
+        self.assertTrue(np.isnan(proj.metas).all())
+
 
 if __name__ == "__main__":
     unittest.main()
diff --git a/Orange/widgets/model/tests/test_owpls.py b/Orange/widgets/model/tests/test_owpls.py
@@ -1,4 +1,5 @@
 import unittest
+import numpy as np
 
 from Orange.data import Table, Domain, StringVariable
 from Orange.widgets.model.owpls import OWPLS
@@ -87,6 +88,21 @@ def test_output_components_multi_target(self):
         self.assertEqual(components.Y.shape, (2, 2))
         self.assertEqual(components.metas.shape, (2, 1))
 
+    def test_missing_target(self):
+        data = self._data[:5].copy()
+        data.Y[[0, 4]] = np.nan
+        self.send_signal(self.widget.Inputs.data, data)
+        output = self.get_output(self.widget.Outputs.data)
+        self.assertFalse(np.isnan(output.metas[:, 3:].astype(float)).any())
+        self.assertTrue(np.isnan(output.metas[0, 1:3].astype(float)).all())
+        self.assertTrue(np.isnan(output.metas[4, 1:3].astype(float)).all())
+        self.assertFalse(np.isnan(output.metas[1:4, 1:3].astype(float)).any())
+
+        with data.unlocked(data.Y):
+            data.Y[:] = np.nan
+        self.send_signal(self.widget.Inputs.data, data)
+        self.assertIsNone(self.get_output(self.widget.Outputs.data))
+
 
 if __name__ == "__main__":
     unittest.main()