simple clusters

clustering.py

#FOR DBSCAN
from sklearn.cluster import DBSCAN
from sklearn import metrics
from sklearn.datasets.samples_generator import make_blobs
from sklearn.preprocessing import StandardScaler
from random import randrange
import math
from numpy import array, argwhere

#For Kmeans
from numpy import ndarray, array
from scipy.cluster import vq


def kmeans_cluster(data, num_clusters):
    features = array(data)
    whitened = vq.whiten(features)
    #book = array((whitened[0],whitened[2]))
    centers, _ = vq.kmeans(whitened, num_clusters)
    cluster, _ = vq.vq(whitened, centers)

    clusters = {}
    for data_idx, cluster_idx in enumerate(cluster):
        items = clusters.get(cluster_idx, [])
        items.append(features[data_idx])
        clusters[cluster_idx] = items

    return clusters.values()


def dbscan_cluster(data):
    data = [[d, 0] for d in data]
    X = StandardScaler().fit_transform(data)

    db = DBSCAN(eps=0.3, min_samples=3).fit(X)
    core_samples = db.core_sample_indices_
    labels = db.labels_

    n_clusters_ = len(set(labels)) - (1 if -1 in labels else 0)
    clusters = [[] for i in range(len(set(labels)))]
    for cluster_idx, l in enumerate(set(labels)):
        members = [idx[0] for idx in argwhere(labels == l)]
        samples = [idx for idx in  core_samples if labels[idx] == l]
        for i in members:
            x = data[i]
            core = i in samples and l != -1
            #clusters[cluster_idx].append([x, core, l])
            clusters[cluster_idx].append(x[0])
    return clusters


if __name__ == '__main__':
    '''ex:
    [4, 6, 6, 6, 1, 21, 23, 22, 22, 28, 48, 46, 48, 45, 48]
    [[21.0, 23.0, 22.0, 22.0, 28.0], [48.0, 46.0, 48.0, 45.0, 48.0], [4.0, 6.0, 6.0, 6.0, 1.0]]
    '''
    
    data = [randrange(1, 10) for i in range(5)]
    data += [randrange(20, 30) for i in range(5)]
    data += [randrange(40, 50) for i in range(5)]

    cluster = dbscan_cluster
    print data
    print cluster(data)