-
Notifications
You must be signed in to change notification settings - Fork 12
/
fishers_method.py
53 lines (46 loc) · 1.22 KB
/
fishers_method.py
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
import numpy as np
def fishers(ip, weights, classes):
ip = np.array(ip)
N, D = ip.shape
weights = np.array(weights)
m1 = []
m2 = []
for idx in range(N):
if classes[idx] == 1:
m1.append(ip[idx])
else:
m2.append(ip[idx])
m1 = np.mean(m1, axis=0)
m2 = np.mean(m2, axis=0)
# between cluster distance
sb = []
sw = []
for w in (weights):
d = (w @ (m1-m2)) ** 2
sb.append(d)
# calculate within cluster distance
sw = []
for w in weights:
running_sw = 0
for idx in range(len(ip)):
if classes[idx] == 1:
running_sw += (w.T @ (ip[idx] - m1)) ** 2
elif classes[idx] == 2:
running_sw += (w.T @ (ip[idx] - m2)) ** 2
sw.append(running_sw)
# print(running_sw)
print("SB: ")
print(sb)
print("SW: ")
print(sw)
cost = []
for _sb, _sw in zip(sb, sw):
cost.append(_sb/_sw)
print("Cost: ")
print(cost)
print("-"*100)
print(f"{weights[np.argmax(cost)]} has high PROJECTION COST")
ip = [[1, 2], [2, 1], [3, 3], [6, 5], [7, 8]]
classes = [1, 1, 1, 2, 2]
weights = [[-1, 5], [2, -3]]
fishers(ip, weights, classes)