matma-proj-11/clustering/kmeans.py

from typing import List

import numpy as np
from cluster import Cluster


class KMeans:
    def __init__(self, n_clusters: int, random_state: int = 0) -> None:
        self.n_clusters = n_clusters
        self.random_state = random_state
        self.clusters: List[Cluster] = list()
        self.labels: np.ndarray = None

    def fit(self, X: np.ndarray) -> None:
        self.__generate_clusters(X)
        # self.labels = np.zeros(X.shape[0])

        # while True:
        #     new_labels = np.array([np.argmin(np.linalg.norm(X - centroid, axis=1)) for centroid in self.centroids])
        #     if np.array_equal(self.labels, new_labels):
        #         break
        #     self.labels = new_labels

        #     for i in range(self.n_clusters):
        #         self.centroids[i] = np.mean(X[self.labels == i], axis=0)

    def __generate_clusters(self, X: np.ndarray) -> None:
        generator = np.random.default_rng(self.random_state)
        min_val, max_val = np.min(X), np.max(X)
        self.clusters = [Cluster(
            generator.uniform(min_val, max_val, X.shape[1])
        ) for _ in range(self.n_clusters)]


# TODO: Delete code below
kmeans = KMeans(2)
kmeans.fit(np.array([[1, 2], [3, 4], [5, 6], [7, 8]]))
for cluster in kmeans.clusters:
    print(cluster.centroid)