machine_learning.k_medoids

k-Medoids Clustering Algorithm

For more details, see: https://en.wikipedia.org/wiki/K-medoids

Functions

_assign_clusters(→ numpy.ndarray)	Assign each data point to the nearest medoid.
_compute_distances(→ numpy.ndarray)	Compute pairwise distances between points and medoids.
_get_data(→ tuple[numpy.ndarray, numpy.ndarray])	Load the Iris dataset and return features and labels.
_initialize_medoids(→ numpy.ndarray)	Randomly select initial medoids.
_update_medoids(→ numpy.ndarray)	Update medoids by minimizing intra-cluster distances.
apply_k_medoids(→ tuple[numpy.ndarray, numpy.ndarray])	Apply k-Medoids clustering to a dataset.
main(→ None)	Run k-Medoids on the Iris dataset and display results.

Module Contents

machine_learning.k_medoids._assign_clusters(distances: numpy.ndarray) → numpy.ndarray

Assign each data point to the nearest medoid.

Args:

distances: Pairwise distance matrix.

Returns:

ndarray: Cluster assignments.

>>> d = np.array([[0.1, 0.4], [0.2, 0.3], [0.9, 0.1]])
>>> _assign_clusters(d)
array([0, 0, 1])

machine_learning.k_medoids._compute_distances(data_matrix: numpy.ndarray, medoids: numpy.ndarray) → numpy.ndarray

Compute pairwise distances between points and medoids.

Args:

data_matrix: Input dataset. medoids: Indices of current medoids.

Returns:

ndarray: Distance matrix of shape (n_samples, n_clusters).

>>> x = np.array([[0.0, 0.0], [1.0, 0.0], [0.0, 1.0]])
>>> d = _compute_distances(x, np.array([0, 2]))
>>> d.shape
(3, 2)

machine_learning.k_medoids._get_data() → tuple[numpy.ndarray, numpy.ndarray]

Load the Iris dataset and return features and labels.

Returns:

tuple[ndarray, ndarray]: Feature matrix and target labels.

>>> features, labels = _get_data()
>>> features.shape
(150, 4)
>>> labels.shape
(150,)

machine_learning.k_medoids._initialize_medoids(n_samples: int, n_clusters: int, random_state: int | None = None) → numpy.ndarray

Randomly select initial medoids.

Args:

n_samples: Total number of samples. n_clusters: Number of clusters. random_state: Optional random seed.

Returns:

ndarray: Indices of initial medoids.

>>> np.random.seed(42)
>>> _initialize_medoids(10, 3).shape
(3,)

machine_learning.k_medoids._update_medoids(data_matrix: numpy.ndarray, clusters: numpy.ndarray, n_clusters: int) → numpy.ndarray

Update medoids by minimizing intra-cluster distances.

Args:

data_matrix: Dataset. clusters: Cluster assignments. n_clusters: Number of clusters.

Returns:

ndarray: Updated medoid indices.

>>> x = np.array([[0.0, 0.0], [1.0, 0.0], [5.0, 0.0]])
>>> clusters = np.array([0, 0, 1])
>>> _update_medoids(x, clusters, 2).shape
(2,)

machine_learning.k_medoids.apply_k_medoids(data_matrix: numpy.ndarray, n_clusters: int = 3, max_iter: int = 100, random_state: int | None = None) → tuple[numpy.ndarray, numpy.ndarray]

Apply k-Medoids clustering to a dataset.

Args:

data_matrix: Input dataset. n_clusters: Number of clusters. max_iter: Maximum iterations. random_state: Optional random seed.

Returns:

tuple[ndarray, ndarray]: Final medoids and cluster assignments.

>>> features, _ = _get_data()
>>> medoids, clusters = apply_k_medoids(features, n_clusters=3, max_iter=10)
>>> len(medoids)
3

machine_learning.k_medoids.main() → None

Run k-Medoids on the Iris dataset and display results.

>>> main()
k-Medoids clustering (first 10 assignments):
[...]