forked from csuldw/MachineLearning
-
Notifications
You must be signed in to change notification settings - Fork 0
Expand file tree
/
Copy pathHandWriting.py
More file actions
61 lines (55 loc) · 1.63 KB
/
Copy pathHandWriting.py
File metadata and controls
61 lines (55 loc) · 1.63 KB
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
# -*- coding: utf-8 -*-
"""
Created on Mon Feb 22 14:41:04 2016
Hand Writing
@author: liudiwei
"""
import os
import numpy as np
import KNN as knn
def imgToVector(filename):
returnVect = np.zeros((1, 1024))
fo = open(filename, 'r')
fr = fo.readlines()
m = len(fr)
for i in range(m):
oneline = fr[i].strip()
n = len(oneline)
if not n:
continue
for j in range(n):
returnVect[0,n*i+j] = int(oneline[j]) #m行n列 当前元素为n*i+j
return returnVect
#加载数据集
def loadDataset(filedir):
trainlist = os.listdir(filedir)
m = len(trainlist)
data_X = np.zeros((m,1024))
data_y = []
for i in range(len(trainlist)):
filename = trainlist[i]
label_i = int(filename.split('_')[0])
data_y.append(label_i)
oneSample_X = imgToVector(filedir + "/" + trainlist[i])
data_X[i,:] = oneSample_X
return data_X, data_y
#调用KNN.py中的KNN算法，并统计数据的错误率
def handwritingClassTest(train_X, train_y, test_X, test_y):
clf = knn.KNNClassifier(k=3)
len_test = len(test_y)
error_count = 0.0
results = clf.classify(test_X, train_X, train_y)
arr_res = np.array(results) - np.array(test_y)
for elem in arr_res:
error_count += np.abs(elem)
error_rate = error_count/len_test
print error_rate
return error_rate
def main():
traindir = "data/trainingDigits"
testdir = "data/testDigits"
train_X, train_y = loadDataset(traindir)
test_X, test_y = loadDataset(testdir)
handwritingClassTest(train_X, train_y, test_X, test_y)
if __name__=="__main__":
main()