updating actionflow

.github/workflows/python-package.yml

@@ -39,4 +39,4 @@ jobs:
         flake8 . --count --exit-zero --max-complexity=10 --max-line-length=127 --statistics
     - name: Test with pytest
       run: |
-        pytest tests/testing_sdtw.py
+        pytest  softdtwkeras/tests/testing_sdtw.py

tests/testing_sdtw.py → softdtwkeras/tests/testing_sdtw.py

@@ -1,30 +1,33 @@
-import numpy as np
-from tslearn.metrics import SquaredEuclidean, SoftDTW
-import tensorflow as tf
-from softdtwkeras.SDTWLoss import SDTWLoss
-
-if __name__ == '__main__':
-    x_ = np.array([[0,0], [0.9,0], [0,0], [0.5,1], [0,0], [0,0]])
-    y_ = np.array([[0,0], [0,0], [0.1,0], [0.9,1], [0,0], [0.5,0]])
-
-
-    x_tf = tf.convert_to_tensor(x_, dtype=tf.float64)
-    y_tf = tf.convert_to_tensor(y_, dtype=tf.float64)
-
-
-    D = SquaredEuclidean(x_, y_)
-    sdtw = SoftDTW(D, gamma=0.001)
-    # soft-DTW discrepancy, approaches DTW as gamma -> 0
-    value = sdtw.compute()
-    # gradient w.r.t. D, shape = [m, n], which is also the expected alignment matrix
-    E = sdtw.grad()
-    # gradient w.r.t. X, shape = [m, d]
-    G = D.jacobian_product(E)
-
-    tff_sdtw__= SDTWLoss(gamma=0.001)
-    val_tensor = tff_sdtw__.unit_loss(x_tf, y_tf)
-
-
-
-    print(value)
-    print(val_tensor)
+import numpy as np
+from tslearn.metrics import SquaredEuclidean, SoftDTW
+import tensorflow as tf
+from softdtwkeras.SDTWLoss import SDTWLoss
+
+def test_soft_dtw():
+    x_ = np.array([[0, 0], [0.9, 0], [0, 0], [0.5, 1], [0, 0], [0, 0]])
+    y_ = np.array([[0, 0], [0, 0], [0.1, 0], [0.9, 1], [0, 0], [0.5, 0]])
+
+    x_tf = tf.convert_to_tensor(x_, dtype=tf.float64)
+    y_tf = tf.convert_to_tensor(y_, dtype=tf.float64)
+
+    D = SquaredEuclidean(x_, y_)
+    sdtw = SoftDTW(D, gamma=0.001)
+    # soft-DTW discrepancy, approaches DTW as gamma -> 0
+    value = sdtw.compute()
+    # gradient w.r.t. D, shape = [m, n], which is also the expected alignment matrix
+    E = sdtw.grad()
+    # gradient w.r.t. X, shape = [m, d]
+    G = D.jacobian_product(E)
+
+    tff_sdtw__ = SDTWLoss(gamma=0.001)
+    val_tensor = tff_sdtw__.unit_loss(x_tf, y_tf)
+    print(value)
+    print(val_tensor)
+    assert np.round(value, decimals=5) == tf.round(val_tensor * 1e5) / 1e5
+
+if __name__ == '__main__':
+    test_soft_dtw()
+
+
+
+