dcpython/meanshift.py at main · ganml/dcpython · GitHub

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
# -*- coding: utf-8 -*-
"""
Created on Sat Feb  1 20:40:38 2025

@author: gjgan
"""

import numpy as np
import matplotlib.pyplot as plt
from scipy.spatial.distance import pdist
from sklearn.datasets import make_blobs, make_circles
from sklearn.metrics import pairwise_distances
from sklearn.cluster import MeanShift
from ucimlrepo import fetch_ucirepo
from dcutil import createCM

def msparabolic(X, sigma=1, tol=1e-8, maxit=300):
    X = np.ascontiguousarray(X)
    Z = X.copy()
    iters = 1
    while iters < maxit:
        W = (pairwise_distances(Z, X) < sigma).astype(float)
        E = np.sum(W, axis=1, keepdims=True)
        Z_ = Z
        Z = (W/E) @ X
        iters += 1
        if np.max(np.abs(Z-Z_)) < tol:
            break
    n, d = X.shape
    clusterCenter = np.zeros((0,d))
    sind = np.argsort(E[:,0])[::-1]
    unassigned = [index.item() for index in sind if index > 0]
    while len(unassigned) > 0:
        i = unassigned[0]
        dist = pairwise_distances(Z[i,:].reshape(1, -1), Z[unassigned,:])
        ind = np.where(dist[0,:] < sigma)[0]
        clusterCenter = np.append(clusterCenter, Z[i,:].reshape(1,-1), axis=0)
        for j in ind:
            sind[sind==unassigned[j]] = -1
        unassigned = [index.item() for index in sind if index > 0]
    dist = pairwise_distances(X, clusterCenter)
    clusterMembership = np.argmin(dist, axis=1)
    return clusterMembership, clusterCenter, Z, iters

def msgaussian(X, sigma=1, tol=1e-8, maxit=300):
    X = np.ascontiguousarray(X)
    Z = X.copy()
    iters = 1
    while iters < maxit:
        W = np.exp(-0.5*np.square(pairwise_distances(Z, X)/sigma))
        E = np.sum(W, axis=1, keepdims=True)
        Z_ = Z
        Z = (W/E) @ X
        iters += 1
        if np.max(np.abs(Z-Z_)) < tol:
            break
    n, d = X.shape
    clusterCenter = np.zeros((0,d))
    sind = np.argsort(E[:,0])[::-1]
    unassigned = [index.item() for index in sind if index > 0]
    while len(unassigned) > 0:
        i = unassigned[0]
        dist = pairwise_distances(Z[i,:].reshape(1, -1), Z[unassigned,:])
        ind = np.where(dist[0,:] < sigma)[0]
        clusterCenter = np.append(clusterCenter, Z[i,:].reshape(1,-1), axis=0)
        for j in ind:
            sind[sind==unassigned[j]] = -1
        unassigned = [index.item() for index in sind if index > 0]
    dist = pairwise_distances(X, clusterCenter)
    clusterMembership = np.argmin(dist, axis=1)
    return clusterMembership, clusterCenter, Z, iters

# synthetic data 1
centers = [[3, 3], [-3, -3], [3, -3]]
X, y = make_blobs(n_samples=300, centers=centers, cluster_std=1, random_state=1)

dm = pdist(X)
sigma = np.percentile(dm, 25)

yhat, cc, Z, iters = msparabolic(X, sigma=sigma)
cm1 = createCM(y, yhat)
print(cm1)

fig, ax = plt.subplots(1, 1, figsize=(6, 6))
ax.scatter(X[:,0], X[:,1], color="white")
for i in range(cc.shape[0]):
    ax.plot(cc[i,0], cc[i,1], "^", markerfacecolor="white",
        markeredgecolor="black", markersize=20, markeredgewidth=2)
for i in range(X.shape[0]):
    ax.text(X[i,0], X[i, 1], str(yhat[i]), fontsize=12)
fig.savefig("blobms1.pdf", bbox_inches='tight')

clustering = MeanShift(bandwidth=sigma).fit(X)
cm = createCM(y, clustering.labels_)
print(cm)

yhat, cc, Z, iters = msgaussian(X, sigma=sigma)
cm1 = createCM(y, yhat)
print(cm1)


# circles

X, y = make_circles(n_samples=300, noise=0.05, factor=0.4, random_state=0)
X = np.round(X, decimals=4)

dm = pdist(X)
sigma = np.percentile(dm, 15)

yhat, cc, Z, iters = msparabolic(X, sigma=sigma)
cm1 = createCM(y, yhat)
print(cm1)

fig, ax = plt.subplots(1, 1, figsize=(6, 6))
ax.scatter(X[:,0], X[:,1], color="white")
for i in range(cc.shape[0]):
    ax.plot(cc[i,0], cc[i,1], "^", markerfacecolor="white",
        markeredgecolor="black", markersize=20, markeredgewidth=2)
for i in range(X.shape[0]):
    ax.text(X[i,0], X[i, 1], str(yhat[i]), fontsize=12)
fig.savefig("circlems1.pdf", bbox_inches='tight')

clustering = MeanShift(bandwidth=sigma).fit(X)
cm = createCM(y, clustering.labels_)
print(cm)

yhat, cc, Z, iters = msgaussian(X, sigma=sigma)
cm1 = createCM(y, yhat)
print(cm1)
yhat, cc, Z, iters = msgaussian(X, sigma=sigma/2)
cm1 = createCM(y, yhat)
print(cm1)

fig, ax = plt.subplots(1, 1, figsize=(6, 6))
ax.scatter(X[:,0], X[:,1], color="white")
for i in range(cc.shape[0]):
    ax.plot(cc[i,0], cc[i,1], "^", markerfacecolor="white",
        markeredgecolor="black", markersize=20, markeredgewidth=2)
for i in range(X.shape[0]):
    ax.text(X[i,0], X[i, 1], str(yhat[i]), fontsize=12)
fig.savefig("circlems2.pdf", bbox_inches='tight')

# iris data

iris = fetch_ucirepo(id=53)
X = iris.data.features
y = iris.data.targets

dist = pdist(X)
sigma = np.percentile(dist, 15)
yhat, cc, Z, iters = msparabolic(X, sigma=sigma)
cm1 = createCM(y, yhat)
print(cm1)

clustering = MeanShift(bandwidth=sigma).fit(X)
cm = createCM(y, clustering.labels_)
print(cm)


yhat, cc, Z, iters = msgaussian(X, sigma=sigma)
cm1 = createCM(y, yhat)
print(cm1)
yhat, cc, Z, iters = msgaussian(X, sigma=sigma/2)
cm1 = createCM(y, yhat)
print(cm1)