Implementation:Scikit learn Scikit learn KNeighborsGraph

Knowledge Sources	Scikit_learn Scikit-learn Docs
Domains	Machine Learning, Graph Construction
Last Updated	2026-02-08 15:00 GMT

Overview

Concrete tool for computing k-nearest neighbors graphs and transformer classes provided by scikit-learn.

Description

This module provides the kneighbors_graph function and the KNeighborsTransformer and RadiusNeighborsTransformer classes for computing weighted graphs of nearest neighbors. The kneighbors_graph function computes a sparse graph representing the k-nearest neighbors connectivity or distance for each point. The transformer classes wrap this functionality into scikit-learn's transformer API, producing sparse matrices suitable for pipeline integration.

Usage

Use kneighbors_graph or KNeighborsTransformer when you need to construct a nearest neighbor graph for downstream tasks such as spectral clustering, manifold learning, or graph-based semi-supervised learning.

Code Reference

Source Location

• Repository: scikit-learn
• File: sklearn/neighbors/_graph.py

Signature

def kneighbors_graph(
    X,
    n_neighbors,
    *,
    mode="connectivity",
    metric="minkowski",
    p=2,
    metric_params=None,
    include_self=False,
    n_jobs=None,
):

class KNeighborsTransformer(
    ClassNamePrefixFeaturesOutMixin, KNeighborsMixin, TransformerMixin, NeighborsBase
):
    def __init__(
        self,
        *,
        mode="distance",
        n_neighbors=5,
        algorithm="auto",
        leaf_size=30,
        metric="minkowski",
        p=2,
        metric_params=None,
        n_jobs=None,
    ):

Import

from sklearn.neighbors import kneighbors_graph, KNeighborsTransformer

I/O Contract

Inputs

Name	Type	Required	Description
X	array-like or sparse matrix	Yes	Input data of shape (n_samples, n_features) or a fitted NearestNeighbors object
n_neighbors	int	Yes	Number of neighbors for each sample
mode	str	No	Type of returned matrix: 'connectivity' or 'distance' (default='connectivity')
metric	str or callable	No	Distance metric to use (default='minkowski')
p	float	No	Power parameter for Minkowski metric (default=2)
metric_params	dict or None	No	Additional keyword arguments for the metric function
include_self	bool or str	No	Whether to mark each sample as its own neighbor (default=False)
n_jobs	int or None	No	Number of parallel jobs

Outputs

Name	Type	Description
A	sparse matrix of shape (n_samples, n_samples)	Graph where A[i, j] gives the weight of the edge connecting i to j (connectivity=1 or distance value)

Usage Examples

Basic Usage

from sklearn.neighbors import kneighbors_graph
import numpy as np

X = np.array([[0, 0], [1, 1], [2, 2], [3, 3]])
A = kneighbors_graph(X, n_neighbors=2, mode="connectivity", include_self=False)
print(A.toarray())

Related Pages

• Principle:Scikit_learn_Scikit_learn_Nearest_Neighbors