Replaced for loop with vectorised mapping on ClusterOverlap regularization layer

tests/test_model_utils.py

@@ -51,14 +51,41 @@ def test_compute_shannon_entropy(tensor):
     tensor=arrays(
         shape=[100, 10],
         dtype=float,
-        unique=False,
+        unique=True,
         elements=st.floats(min_value=0.0, max_value=10.0),
-    )
+    ),
+    k=st.integers(min_value=5, max_value=20),
 )
 def test_k_nearest_neighbors(tensor, k):
+
     deepof_knn = deepof.model_utils.get_k_nearest_neighbors(tensor, k, 0)
-    sklearn_knn = NearestNeighbors().fit(tensor)
-    assert np.allclose(deepof_knn, sklearn_knn.kneighbors())
+    sklearn_knn = NearestNeighbors(k).fit(tensor)
+    sklearn_knn = sklearn_knn.kneighbors(tensor[0].reshape(1, -1))[1].flatten()
+    assert np.allclose(deepof_knn.numpy(), sorted(sklearn_knn))
+
+
+@settings(deadline=None, suppress_health_check=[HealthCheck.too_slow])
+@given(
+    tensor=arrays(
+        shape=[100, 10],
+        dtype=float,
+        unique=True,
+        elements=st.floats(min_value=0.0, max_value=10.0),
+    ),
+    clusters=arrays(
+        shape=[100],
+        dtype=int,
+        unique=False,
+        elements=st.integers(min_value=0, max_value=10),
+    ),
+    k=st.integers(min_value=5, max_value=20),
+)
+def test_get_neighbourhood_entropy(tensor, clusters, k):
+
+    neighborhood_entropy = deepof.model_utils.get_neighbourhood_entropy(
+        0, tensor, clusters, k
+    ).numpy()
+    assert isinstance(neighborhood_entropy, np.float32)
 
 
 @settings(deadline=None, suppress_health_check=[HealthCheck.too_slow])