Statistics.nanunique: Pass all parameters to unique function call

Author: pavlin-policar

Orange/statistics/util.py

@@ -441,11 +441,17 @@ def unique(x, return_counts=False):
         return np.insert(r, zero_index, 0)
 
 
-def nanunique(x):
+def nanunique(*args, **kwargs):
     """ Return unique values while disregarding missing (np.nan) values.
     Supports sparse or dense matrices. """
-    r = unique(x)
-    return r[~np.isnan(r)]
+    result = unique(*args, **kwargs)
+
+    if isinstance(result, tuple):
+        result, counts = result
+        non_nan_mask = ~np.isnan(result)
+        return result[non_nan_mask], counts[non_nan_mask]
+    
+    return result[~np.isnan(result)]
 
 
 def digitize(x, bins, right=False):

Orange/tests/test_statistics.py

@@ -115,13 +115,6 @@ def test_nanmin_nanmax(self):
                     nanmax(X_sparse, axis=axis),
                     np.nanmax(X, axis=axis))
 
-    def test_nanunique(self):
-        x = csr_matrix(np.array([0, 1, 1, np.nan]))
-        np.testing.assert_array_equal(
-            nanunique(x),
-            np.array([0, 1])
-        )
-
     def test_mean(self):
         for X in self.data:
             X_sparse = csr_matrix(X)
@@ -400,3 +393,21 @@ def test_sparse_explicit_zeros(self):
             unique(y, return_counts=True),
             unique(x, return_counts=True),
         )
+
+    @dense_sparse
+    def test_nanunique_ignores_nans_in_values(self, array):
+        x = array([[-1., 1., 0., 2., 3., np.nan],
+                   [ 0., 0., 0., 3., 5., np.nan],
+                   [-1., 0., 0., 1., 7.,     6.]])
+        expected = [-1, 0, 1, 2, 3, 5, 6, 7]
+
+        np.testing.assert_equal(nanunique(x, return_counts=False), expected)
+
+    @dense_sparse
+    def test_nanunique_ignores_nans_in_counts(self, array):
+        x = array([[-1., 1., 0., 2., 3., np.nan],
+                   [ 0., 0., 0., 3., 5., np.nan],
+                   [-1., 0., 0., 1., 7.,     6.]])
+        expected = [2, 6, 2, 1, 2, 1, 1, 1]
+
+        np.testing.assert_equal(nanunique(x, return_counts=True)[1], expected)