Merge fd89c66 into 2bc7fce

kmodes/util/dissim.py

@@ -56,3 +56,18 @@ def calc_dissim(b, X, memj, idr):
                                if b[idr] == t else 1.0
                                for idr, t in enumerate(val_a)]).sum(0)
                      for idj, val_a in enumerate(a)])
+
+
+def jaccard_binary_dissim(a, b, **_):
+    """Jaccard distance function"""
+
+    if np.isnan(a).any() or np.isnan(b).any():
+        raise ValueError("Missing values detected in numerical columns.")
+
+    i = np.sum((a == 1) & (b == 1), axis=1)
+    j = np.sum((a == 1) | (b == 1), axis=1)
+
+    if j != 0:
+        return ((j - i) / j)
+    else:
+        return 1

kmodes/util/tests/test_dissim.py

@@ -8,6 +8,7 @@
 from sklearn.utils.testing import assert_equal, assert_array_equal
 
 from kmodes.util.dissim import matching_dissim, euclidean_dissim, ng_dissim
+from kmodes.util.dissim import jaccard_binary_dissim
 
 
 class TestDissimilarityMeasures(unittest.TestCase):
@@ -103,3 +104,21 @@ def test_ng_dissim(self):
                            ng_dissim(centroids, X[1], X=X, membship=membship))
         assert_array_equal(np.array([mdiss_12, mdiss_22]),
                            ng_dissim(centroids, X[2], X=X, membship=membship))
+
+    def test_jaccard__binary_dissim(self):
+        a = np.array([[1, 0, 0, 1, 1, 0]])
+        b = np.array([[1, 1, 1, 0, 1, 1]])
+        assert_equal(2 / 3, jaccard_binary_dissim(a, b))
+
+        a = np.array([[0, 0, 1, 1, 0, 0]])
+        b = np.array([[0, 1, 1, 0, 0, 1]])
+        assert_equal(0.75, jaccard_binary_dissim(a, b))
+
+        a = np.array([[0, 0, 0, 0, 0, 0]])
+        b = np.array([[0, 0, 0, 0, 0, 0]])
+        assert_equal(1, jaccard_binary_dissim(a, b))
+
+        a = np.array([[np.NaN, 0, 0, 1, 1, 0]])
+        b = np.array([[1, 1, 1, 0, 1, 1]])
+        with self.assertRaises(ValueError):
+            jaccard_binary_dissim(a, b)