vector_array.append(vector_create(line[1],word_r,word_n))
pass
for line in lines_tr:
line = line.split('\t',2)
train_array.append(vector_create(line[2],word_r,word_n))
flag.append(line[1])
pass
#svm
clf = modle(train_array,flag)
print "predicting"
fileHandler_out = open ('data\\result.csv',"w")
num = 800001
for pr in vector_array:
tag = []
tag.append(str(num))
tag.append(predict(clf,pr))
result.append(tag)
num += 1
pass
#change the result
# fileHandler = open('data\\pat1_2_3.3')
# fileHandler.seek(0)
# ch = fileHandler.readlines()
# for x in ch:
# for r in result:
# if x == r[0]:
# r[1]=1
# pass
# pass
# pass
for r in result:
result = []
word_r, word_n = fileUtil.read_dict()
print 'set train vector'
for list1 in list_train:
train_array.append(vector_create(list1[2], word_r, word_n))
flag.append(list1[1])
pass
print 'set test vector'
for list2 in list_test:
test_array.append(vector_create(list2[2], word_r, word_n))
answer.append(list2[1])
pass
print "read finished"
# svm
clf = modle(train_array, flag)
result = predict(clf, test_array)
sum_rubbish = 0
sum_normal = 0
pre_rubbish = 0
pre_normal = 0
right_rubbish = 0
right_normal = 0
r_pre_n = []
for x in xrange(0, len(result)-1):
if result[x] == answer[x]:
if result[x] == "1":
sum_rubbish += 1
pre_rubbish += 1
right_rubbish += 1
else:
sum_normal += 1
