scikit-learn
diff --git a/‎doc/whats_new/v1.1.rst
Lines changed: 5 additions & 0 deletions b/‎doc/whats_new/v1.1.rst
Lines changed: 5 additions & 0 deletions
diff --git a/‎sklearn/neighbors/_regression.py
Lines changed: 1 addition & 1 deletion b/‎sklearn/neighbors/_regression.py
Lines changed: 1 addition & 1 deletion
@@ -664,6 +664,11 @@ Changelog
   instead of `__init__`. :pr:`21430` by :user:`Desislava Vasileva <DessyVV>` and
   :user:`Lucy Jimenez <LucyJimenez>`.
 
+- |Fix| :func:`neighbors.KNeighborsRegressor.predict` now works properly when
+  given an array-like input if `KNeighborsRegressor` is first constructed with a
+  callable passed to the `weights` parameter. :pr:`22687` by
+  :user:`Meekail Zain <micky774>`
+
 :mod:`sklearn.neural_network`
 .............................
 
 
@@ -233,7 +233,7 @@ def predict(self, X):
         if weights is None:
             y_pred = np.mean(_y[neigh_ind], axis=1)
         else:
-            y_pred = np.empty((X.shape[0], _y.shape[1]), dtype=np.float64)
+            y_pred = np.empty((neigh_dist.shape[0], _y.shape[1]), dtype=np.float64)
             denom = np.sum(weights, axis=1)
 
             for j in range(_y.shape[1]):
 
@@ -30,7 +30,10 @@
 )
 from sklearn.model_selection import cross_val_score
 from sklearn.model_selection import train_test_split
-from sklearn.neighbors import VALID_METRICS_SPARSE
+from sklearn.neighbors import (
+    VALID_METRICS_SPARSE,
+    KNeighborsRegressor,
+)
 from sklearn.neighbors._base import (
     _is_sorted_by_data,
     _check_precomputed,
@@ -2096,3 +2099,19 @@ def test_radius_neighbors_brute_backend(
 def test_valid_metrics_has_no_duplicate():
     for val in neighbors.VALID_METRICS.values():
         assert len(val) == len(set(val))
+
+
+def test_regressor_predict_on_arraylikes():
+    """Ensures that `predict` works for array-likes when `weights` is a callable.
+
+    Non-regression test for #22687.
+    """
+    X = [[5, 1], [3, 1], [4, 3], [0, 3]]
+    y = [2, 3, 5, 6]
+
+    def _weights(dist):
+        return np.ones_like(dist)
+
+    est = KNeighborsRegressor(n_neighbors=1, algorithm="brute", weights=_weights)
+    est.fit(X, y)
+    assert_allclose(est.predict([[0, 2.5]]), [6])