scikit-learn
diff --git a/‎sklearn/cluster/tests/test_k_means.py
Copy file name to clipboardExpand all lines: sklearn/cluster/tests/test_k_means.py
+3-12Lines changed: 3 additions & 12 deletions b/‎sklearn/cluster/tests/test_k_means.py
Copy file name to clipboardExpand all lines: sklearn/cluster/tests/test_k_means.py
+3-12Lines changed: 3 additions & 12 deletions
@@ -438,32 +438,23 @@ def test_minibatch_sensible_reassign(global_random_seed):
     ).fit(zeroed_X)
     # there should not be too many exact zero cluster centers
     num_non_zero_clusters = km.cluster_centers_.any(axis=1).sum()
-    if num_non_zero_clusters < 10:
-        raise ValueError(
-            f"Number of non-zero clusters is too small {num_non_zero_clusters=}"
-        )
+    assert num_non_zero_clusters > 9, f"{num_non_zero_clusters=} is too small"
 
     # do the same with batch-size > X.shape[0] (regression test)
     km = MiniBatchKMeans(
         n_clusters=20, batch_size=200, random_state=global_random_seed, init="random"
     ).fit(zeroed_X)
     # there should not be too many exact zero cluster centers
     num_non_zero_clusters = km.cluster_centers_.any(axis=1).sum()
-    if num_non_zero_clusters < 10:
-        raise ValueError(
-            f"Number of non-zero clusters is too small {num_non_zero_clusters=}"
-        )
+    assert num_non_zero_clusters > 9, f"{num_non_zero_clusters=} is too small"
 
     # do the same with partial_fit API
     km = MiniBatchKMeans(n_clusters=20, random_state=global_random_seed, init="random")
     for i in range(100):
         km.partial_fit(zeroed_X)
     # there should not be too many exact zero cluster centers
     num_non_zero_clusters = km.cluster_centers_.any(axis=1).sum()
-    if num_non_zero_clusters < 10:
-        raise ValueError(
-            f"Number of non-zero clusters is too small {num_non_zero_clusters=}"
-        )
+    assert num_non_zero_clusters > 9, f"{num_non_zero_clusters=} is too small"
 
 
 @pytest.mark.parametrize(