Adding sbd distance method and its test.

khiva/distances.py

@@ -28,8 +28,7 @@ def euclidean(tss):
             between time series 0 and time series 1.
     """
     b = ctypes.c_void_p(0)
-    KhivaLibrary().c_khiva_library.euclidean(ctypes.pointer(tss.arr_reference),
-                                             ctypes.pointer(b))
+    KhivaLibrary().c_khiva_library.euclidean(ctypes.pointer(tss.arr_reference), ctypes.pointer(b))
     return Array(array_reference=b)
 
 
@@ -43,8 +42,7 @@ def dtw(tss):
             distance between time series 0 and time series 1.
     """
     b = ctypes.c_void_p(0)
-    KhivaLibrary().c_khiva_library.dtw(ctypes.pointer(tss.arr_reference),
-                                       ctypes.pointer(b))
+    KhivaLibrary().c_khiva_library.dtw(ctypes.pointer(tss.arr_reference), ctypes.pointer(b))
     return Array(array_reference=b)
 
 
@@ -58,8 +56,7 @@ def hamming(tss):
             between time series 0 and time series 1.
     """
     b = ctypes.c_void_p(0)
-    KhivaLibrary().c_khiva_library.hamming(ctypes.pointer(tss.arr_reference),
-                                           ctypes.pointer(b))
+    KhivaLibrary().c_khiva_library.hamming(ctypes.pointer(tss.arr_reference), ctypes.pointer(b))
     return Array(array_reference=b)
 
 
@@ -73,8 +70,22 @@ def manhattan(tss):
             between time series 0 and time series 1.
     """
     b = ctypes.c_void_p(0)
-    KhivaLibrary().c_khiva_library.manhattan(ctypes.pointer(tss.arr_reference),
-                                             ctypes.pointer(b))
+    KhivaLibrary().c_khiva_library.manhattan(ctypes.pointer(tss.arr_reference), ctypes.pointer(b))
+    return Array(array_reference=b)
+
+
+def sbd(tss):
+    """ Calculates the Shape-Based distance (SBD). It computes the normalized cross-correlation and
+    it returns the value that maximizes the correlation value between time series.
+
+    :param tss: Expects an input array whose dimension zero is the length of the time series (all the same) and
+                dimension one indicates the number of time series.
+    :return: Array with an upper triangular matrix where each position corresponds to the distance between two time series.
+            Diagonal elements will be zero. For example: Position row 0 column 1 records the distance between time series 0
+            and time series 1.
+    """
+    b = ctypes.c_void_p(0)
+    KhivaLibrary().c_khiva_library.sbd(ctypes.pointer(tss.arr_reference), ctypes.pointer(b))
     return Array(array_reference=b)
 
 
@@ -88,6 +99,5 @@ def squared_euclidean(tss):
             and time series 1.
     """
     b = ctypes.c_void_p(0)
-    KhivaLibrary().c_khiva_library.squared_euclidean(ctypes.pointer(tss.arr_reference),
-                                                     ctypes.pointer(b))
+    KhivaLibrary().c_khiva_library.squared_euclidean(ctypes.pointer(tss.arr_reference), ctypes.pointer(b))
     return Array(array_reference=b)

tests/unit_tests/distances_unit_tests.py

@@ -31,12 +31,6 @@ def test_euclidean(self):
         expected = np.array([0, 0, 0, 8, 0, 0, 16, 8, 0])
         np.testing.assert_array_almost_equal(euclidean_result, expected, decimal=1)
 
-    def test_squared_euclidean(self):
-        squared_euclidean_result = squared_euclidean(
-            Array(data=[[0, 1, 2, 3], [4, 5, 6, 7], [8, 9, 10, 11]])).to_numpy().flatten()
-        expected = np.array([0, 0, 0, 64, 0, 0, 256, 64, 0])
-        np.testing.assert_array_almost_equal(squared_euclidean_result, expected, decimal=1)
-
     def test_dtw(self):
         euclidean_result = dtw(Array(
             data=[[1, 1, 1, 1, 1], [2, 2, 2, 2, 2], [3, 3, 3, 3, 3], [4, 4, 4, 4, 4], [5, 5, 5, 5, 5]])).to_numpy()
@@ -55,6 +49,18 @@ def test_manhattan(self):
         expected = np.array([[0, 0, 0, 0, 0], [5, 0, 0, 0, 0], [10, 5, 0, 0, 0], [15, 10, 5, 0, 0], [20, 15, 10, 5, 0]])
         np.testing.assert_array_almost_equal(result, expected, decimal=2)
 
+    def test_sbd(self):
+        sbd_result = sbd(
+            Array(data=[[1, 2, 3, 4, 5], [1, 1, 0, 1, 1], [10, 12, 0, 0, 1]])).to_numpy().flatten()
+        expected = np.array([0, 0, 0, 0.505025, 0, 0, 0.458583, 0.564093, 0])
+        np.testing.assert_array_almost_equal(sbd_result, expected, decimal=1)
+
+    def test_squared_euclidean(self):
+        squared_euclidean_result = squared_euclidean(
+            Array(data=[[0, 1, 2, 3], [4, 5, 6, 7], [8, 9, 10, 11]])).to_numpy().flatten()
+        expected = np.array([0, 0, 0, 64, 0, 0, 256, 64, 0])
+        np.testing.assert_array_almost_equal(squared_euclidean_result, expected, decimal=1)
+
 
 if __name__ == '__main__':
     suite = unittest.TestLoader().loadTestsFromTestCase(DistancesTest)