Machine-Learning-Algorithms/hw8.py at master · HarshaliWagh/Machine-Learning-Algorithms · 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
import sys
import random
import math

datafile = sys.argv[1]

f = open(datafile, 'r')
data = []
a = f.readline()
while (a != ''):
    b = a.split()
    a2 = []
    for c in range(0, len(b), 1):
        a2.append(float(b[c]))
    data.append(a2)

    a = f.readline()

rows = len(data)
cols = len(data[0])

f.close()

d = int(sys.argv[2])
col = [0 for _ in range(0, cols, 1)]
e = [col for _ in range(0, d, 1)]
value = 0

for f in range(0, d, 1):
    value = random.randrange(1, rows-1)
    e[f] = data[value]

trainlabels = {}
difference = 1

prev = [[0]*cols for g in range(d)]
mdist = [0 for _ in range(0, d, 1)]

n = [0.1 for _ in range(0, d, 1)]
distance = [0.1 for _ in range(0, d, 1)]

totaldistance = 1
classes = []

while ((totaldistance) > 0):
    for h in range(0,rows, 1):
        distance =[]

        for f in range(0, d, 1):
            distance.append(0)

        for f in range(0, d, 1):
            for c in range(0, cols, 1):
                distance[f] += ((data[h][c] - e[f][c])**2)

        for f in range(0, d, 1):
            distance[f] = (distance[f])**0.5

        mindist=0
        mindist = min(distance)
        for f in range(0, d, 1):
            if(distance[f]==mindist):
                trainlabels[h] = f
                n[f]+=1
                break

    e = [[0]*cols for g in range(d)]
    col = []

    for h in range(0, rows, 1):
        for f in range(0, d, 1):
            if(trainlabels.get(h) == f):
                for c in range(0, cols, 1):
                    temp =  e[f][c]
                    temp1 =  data[h][c]
                    e[f][c] = temp + temp1
    for c in range(0, cols, 1):
        for h in range(0, d, 1):
            e[h][c] = e[h][c]/n[h]
    classes = [int(g) for g in n]
    n=[0.1]*d
    #print("m",e)

    dist = []
    for f in range(0, d, 1):
        dist.append(0)
    for f in range(0, d, 1):
        for c in range(0, cols, 1):
            dist[f]+=float((prev[f][c]-e[f][c])**2)

        dist[f] = (dist[f])**0.5
    prev=e
    totaldistance = 0
    for i in range(0,len(dist),1):
        totaldistance += dist[i]

#print("distance between means:",totaldistance)
#print("data in each cluster for k =",d,"is",classes)

for h in range(0,rows, 1):
    print(trainlabels[h],h)