[BUG] `kneighbors_graph` silently ignores `Minkowski` feature weights in `metric_params`

[apiqwe]

Describe the bug

cuml.neighbors.kneighbors_graph silently ignores the Minkowski feature-weight vector supplied through metric_params={"w": ...}.

For points [0,0] and [1,2], with p=2 and feature weights [1,2], the weighted Minkowski distance is:

(1 * |1|^2 + 2 * |2|^2)^(1/2) = sqrt(9) = 3

Scikit-learn returns 3.0, while cuML returns 2.236068, which is sqrt(5) and exactly matches the unweighted Euclidean distance. No warning or unsupported-parameter error is raised.

Steps/Code to reproduce bug

cuML reproducer:

import numpy as np
from cuml.neighbors import kneighbors_graph 

X = np.array([
    [0., 0.],
    [1., 2.],
])

kwargs = dict(
    n_neighbors=2,
    mode="distance",
    metric="minkowski",
    p=2,
    metric_params={"w": np.array([1.0, 2.0])},
    include_self=True,
)

a = kneighbors_graph(X, **kwargs).toarray()

print(a)

Output:

[[0.       2.236068]
 [2.236068 0.      ]]

For comparison, the equivalent scikit-learn code:

import numpy as np
from sklearn.neighbors import kneighbors_graph 

X = np.array([
    [0., 0.],
    [1., 2.],
])

kwargs = dict(
    n_neighbors=2,
    mode="distance",
    metric="minkowski",
    p=2,
    metric_params={"w": np.array([1.0, 2.0])},
    include_self=True,
)

a = kneighbors_graph(X, **kwargs).toarray()

print(a)

Output:

[[0. 3.]
 [3. 0.]]

Expected behavior

metric_params={"w": np.array([1.0, 2.0])} should affect the Minkowski distance calculation, producing 3.0 for the off-diagonal entries:

[[0. 3.]
 [3. 0.]]

If weighted Minkowski distance is intentionally unsupported, cuML should reject the w parameter with a clear error instead of silently returning an unweighted result.

Environment details (please complete the following information):

• Environment location: Docker
• Linux Distro/Architecture: Ubuntu 24.04 / x86_64
• GPU Model/Driver: NVIDIA GeForce RTX 4090 / 595.71.05
• CUDA: 13.2
• Method of cuDF & cuML install: conda

conda list:

conda list
# packages in environment at /opt/conda/envs/rapids-26.08:
#
# Name                              Version          Build                                         Channel
# Name              Version       Build                                      Channel
python              3.14.6        h242f9ac_102_cp314                         conda-forge
numpy               2.4.6         py314h2b28147_0                            conda-forge
scipy               1.16.3        py314hf07bd8e_2                            conda-forge
scikit-learn        1.9.0         np2py314hf09ca88_0                         conda-forge
rapids              26.08.00      cuda13_260806_c2656556                     rapidsai
cuml                26.08.00      cuda13_cp311_abi3_260805_265b9da6          rapidsai
libcuml             26.08.00      cuda13_260805_265b9da6                     rapidsai
cudf                26.08.00      cuda13_cp311_abi3_260805_ff5b362d          rapidsai
libraft             26.08.00      cuda13_260805_ebf92684                     rapidsai
libraft-headers     26.08.00      cuda13_260805_ebf92684                     rapidsai
pylibraft           26.08.00      cuda13_cp311_abi3_260805_ebf92684          rapidsai
cuvs                26.08.01      cuda13_cp311_abi3_260806_25b1be43          rapidsai
libcuvs             26.08.01      cuda13_260806_25b1be43                     rapidsai
cupy                14.1.1        py314hdea9c46_0                            conda-forge
cupy-core           14.1.1        py314hcd3b49b_0                            conda-forge
numba               0.64.0        py314h8169c2f_0                            conda-forge
numba-cuda          0.30.4        py314h42812f9_0                            conda-forge
rmm                 26.08.00      cuda13_cp311_abi3_260805_42d059f1          rapidsai
librmm              26.08.00      cuda13_260805_42d059f1                     rapidsai
cuda-version        13.3           hcbadf70_3                                 conda-forge
cuda-bindings       13.3.1        py314h42812f9_1                            conda-forge
cuda-cudart         13.3.29       hecca717_0                                 conda-forge
cuda-nvrtc          13.3.33       hecca717_0                                 conda-forge
libcublas           13.6.0.2      h676940d_0                                 conda-forge
libcusolver         12.2.6.9      h676940d_0                                 conda-forge
libcusparse         12.8.2.51     hecca717_0                                 conda-forge
libcurand           10.4.3.29     h676940d_0                                 conda-forge

Additional context

The only difference between the weighted and unweighted calculations is the second feature's weight:

unweighted: sqrt(1^2 + 2^2)         = sqrt(5) = 2.2360679...
weighted:   sqrt(1*1^2 + 2*2^2)     = sqrt(9) = 3.0

The cuML result therefore demonstrates that metric_params["w"] is not applied.

The scikit-learn kneighbors_graph documentation defines metric_params as additional keyword arguments for the metric. Silently ignoring an accepted metric parameter can produce plausible-looking but incorrect graph weights in downstream manifold-learning and clustering algorithms.

[jcrist]

👋 I've noticed this account has opened a large quantity of issues in the last few weeks.

The issues are all mostly valid and have reproducible examples (which is great)!

However, given:

• The large number and high rate of issues opened
• The wide variety of estimators and functions used
• The wide variety of valid and invalid inputs tried

I have to assume you're using a bot/LLM/tool of some kind to find issues in cuml.

cuml and other libraries are maintained by humans. We're very happy to help other humans use our library and resolve bugs found in real-world usage. cuml definitely has bugs. However, we also have only so much capacity to resolve bugs, and would prefer to spend our time fixing bugs that impact real world use cases. If any of these reflect that, then great. But if you're just reporting errors that a bot or parameter sweep finds, then that's less useful for us. Please be respectful of our time - we want to help, but can only do so much. We don't have anything written up in our docs around LLM policies, but if we did it would probably look similar to what cupy has here.

TL;DR: we like user submitted issues, but only if they're related to real work you're doing and not just bughunting.