import numpy as np import math import random def normal mu sigma retur

 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
import numpy as np
import math
import random
def normal(x, mu, sigma):
return (1/(sigma*(2*math.pi)**0.5))*math.exp(-(x - 2.*random.random() - mu))**2/(2*sigma**2)
def normal_distribution(mu, sigma):
return sigma*random.random() + mu
dots = []
x, sigma = 100, 0.5
for mu in [2., 4., 6.]:
dots += [[normal_distribution(mu, sigma),normal_distribution(mu, sigma), normal_distribution(mu, sigma)] for _ in range(3)]
print dots
#print '['
#for [a, b, c] in dots:
# print a ,',', b, ',', c, ';'
#print ']'
def metrics(point_1, point_2):
[x1, y1, z1] = point_1
print 'point_1', point_1
[x2, y2, z2] = point_2
print 'point_2', point_2
a = (x1-x2)**2 + (y1-y2)**2 + (z1-z2)**2
print 'metrix = ', a
return a
a = metrics([1,1,1], [0, 0, 0])
print a
n = 3
centers = 6.0*np.random.random((n, 3))
print centers
def get_nearest(x, centers):
return np.argmin([metrics(x, c) for c in centers])
lables = [get_nearest(dot, centers) for dot in dots]
print 'lables'
print lables
def recalck_center(dots, indexes):
x, y, z = 0., 0., 0.
for i in indexes:
[a, b, c] = dots[i]
x += a
y += b
z += c
return [x / len(indexes), y / len(indexes), z / len(indexes)]
def get_dots_for_labels(klaster_index, lables_for_all):
result = []
i=0
print 'for klaster ', klaster_index
for lable in lables_for_all:
if klaster_index == lable:
result += [i]
i += 1
print 'vot rezultat:'
print result
return result
centers = 6.0*np.random.random((3, 3))
centers = [[2., 2., 2.], [4., 4., 4.], [6., 6., 6.]]
print 'centers:'
print centers
def my_kmeans(dots, klasters_number, iterations=3):
centers = 6.0*np.random.random((klasters_number, 3))
for i in xrange(iterations):
lables = [get_nearest(dot, centers) for dot in dots]
for klaster_index in range(klasters_number):
current_claster_indexes = get_dots_for_labels(klaster_index, lables)
#print current_claster_indexes
#centers[klaster_index] = recalck_center(dots, current_claster_indexes)
return lables
my_kmeans(dots, 3)