normalizer/normalizerOffline.py at master · tinchoa/normalizer · 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
import numpy as np
import scipy.stats as st
import math,time
from sklearn import preprocessing


class normalizer:

	def run(self):
		w = open("classes-17-norm-offline.out",'w')
		f = open("classes-17.out",'r')
		linha = f.readline()
		vetor = [float(i) for i in linha.split(",")[5:-1]]
		#vetor
		vetorMax = vetor
		vetorMin = [float(i) for i in linha.split(",")[5:-1]]
		while linha != "":
			vetor = [float(i) for i in linha.split(",")[5:-1]]
			for i in range(len(vetor)):
				if vetor[i] > vetorMax[i]:
					vetorMax[i] = vetor[i]
				if vetor[i] < vetorMin[i]:
					vetorMin[i] = vetor[i]
			linha = f.readline()

		f = open("classes-17.out",'r')
		linha = f.readline()
		salida=[]
		while linha != "":
			vetor = [float(i) for i in linha.split(",")[5:-1]]
			for i in range(len(vetor)):
				try:
					vetor[i] = (vetor[i]-vetorMin[i])/(vetorMax[i]-vetorMin[i])
				except ZeroDivisionError:
					if (vetorMin[i]+vetorMax[i]) > 0: vetor[i] = (vetor[i])/(vetorMax[i]+vetorMin[i])
					else: vetor[i] = 0.5
			linha = f.readline()
			#armazanar em arquivo
			vetorStr = [str(i) for i in vetor]
			w.write(",".join(vetorStr)+"\n")
			salida.append(vetor)

		for i in range(len(salida)):
			if i==0:
				output=salida[0]
			else:
				output=np.vstack((output,salida[i]))

		return output.tolist()