Implementation:Scikit learn Scikit learn AffinityPropagation

Knowledge Sources	Scikit_learn Scikit-learn Docs
Domains	Clustering, Unsupervised Learning
Last Updated	2026-02-08 15:00 GMT

Overview

Concrete tool for performing affinity propagation clustering provided by scikit-learn.

Description

AffinityPropagation is a clustering algorithm that identifies exemplars among data points and forms clusters by passing messages between data points. Unlike K-Means, it does not require specifying the number of clusters in advance; instead, it determines the number of clusters based on preference values and pairwise similarity. The algorithm iteratively refines responsibility and availability matrices until convergence or until the maximum number of iterations is reached.

Usage

Use AffinityPropagation when you need a clustering algorithm that automatically determines the number of clusters from the data, especially when you have a precomputed similarity matrix or want to identify representative exemplars. It is well-suited for small to medium datasets where the number of clusters is unknown.

Code Reference

Source Location

• Repository: scikit-learn
• File: sklearn/cluster/_affinity_propagation.py

Signature

class AffinityPropagation(ClusterMixin, BaseEstimator):
    def __init__(
        self,
        *,
        damping=0.5,
        max_iter=200,
        convergence_iter=15,
        copy=True,
        preference=None,
        affinity="euclidean",
        verbose=False,
        random_state=None,
    ):

Import

from sklearn.cluster import AffinityPropagation

I/O Contract

Inputs

Name	Type	Required	Description
damping	float	No	Damping factor in the range [0.5, 1.0) controlling message update smoothing. Default is 0.5.
max_iter	int	No	Maximum number of iterations. Default is 200.
convergence_iter	int	No	Number of iterations with no change in cluster count to declare convergence. Default is 15.
copy	bool	No	Whether to make a copy of input data. Default is True.
preference	array-like or float	No	Preferences for each point; higher values increase likelihood of being chosen as exemplar. Default is None (median of similarities).
affinity	str	No	Affinity metric to use: "euclidean" or "precomputed". Default is "euclidean".
verbose	bool	No	Whether to print progress messages. Default is False.
random_state	int or RandomState	No	Random state for reproducibility. Default is None.

Outputs

Name	Type	Description
cluster_centers_indices_	ndarray of shape (n_clusters,)	Indices of cluster centers (exemplars).
cluster_centers_	ndarray of shape (n_clusters, n_features)	Cluster center coordinates.
labels_	ndarray of shape (n_samples,)	Cluster label for each sample.
affinity_matrix_	ndarray of shape (n_samples, n_samples)	Affinity matrix used for clustering.
n_iter_	int	Number of iterations taken to converge.

Usage Examples

Basic Usage

from sklearn.cluster import AffinityPropagation
import numpy as np

X = np.array([[1, 2], [1, 4], [1, 0],
              [10, 2], [10, 4], [10, 0]])

clustering = AffinityPropagation(random_state=5).fit(X)
print(clustering.labels_)
print(clustering.cluster_centers_)

Related Pages

• Principle:Scikit_learn_Scikit_learn_Clustering