-
Notifications
You must be signed in to change notification settings - Fork 0
/
minimizeByResultCovariance.py
54 lines (49 loc) · 1.43 KB
/
minimizeByResultCovariance.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
54
import binarisationVariables as BV
import numpy as np
from removeUselessPoints import *
from xlutils.copy import copy
from xlrd import open_workbook
from xlwt import easyxf
import xlsxwriter
import json
def createMinimisedSupportSetOutput(uselessPoints):
workbook = xlsxwriter.Workbook('minimizedSupportSetOutput.xls')
worksheet = workbook.add_worksheet()
row = 0
temp = 0
uselessPointsSize = len(uselessPoints)
for item in BV.items:
i = 0
column = 0
j = 0
for attribute in item:
if i == BV.numberOfAttributes - 1:
worksheet.write(row, column - j, attribute)
continue
if j < uselessPointsSize and uselessPoints[j] == i:
j = j + 1
elif j == uselessPointsSize:
worksheet.write(row, column - j, attribute)
else:
worksheet.write(row, column - j, attribute)
column = column + 1
i = i + 1
row = row + 1
workbook.close()
def minimizeByResultCovariance(threshold):
uselessPoints = []
correlation = []
i = 0
while i < BV.numberOfAttributes - 1:
correlationCoef = abs(np.nan_to_num((np.corrcoef([row[i] for row in BV.items],[row[BV.numberOfAttributes-1] for row in BV.items])))[0][1])
if correlationCoef < threshold:
uselessPoints.append(i)
else:
correlation.append(correlationCoef)
i = i + 1
BV.uselessPoints = uselessPoints
createMinimisedSupportSetOutput(uselessPoints)
removeUselessPoints()
fileObj = open("finalCorrelationValue.txt","w")
json.dump(correlation, fileObj)
fileObj.close()