features=[] #数据集特征集

labels=[] #数据集类标集

features_t=[]

maxProbability_t=[]

tables_result=[]

Merchant_ids_t=[]

Merchant_ids_test={} #商家ID字典,test

Merchant_ids_train={} #商家ID字典,train

testData=[]

trainData=[]

#测试数据集读取

test_data = open('./test_data/data_revised.csv')

for line in test_data.readlines():

lineArr = line.strip().split(',')

Merchant_ids_t.append(int(lineArr[1]))

features_t.append([float(lineArr[3]), int(lineArr[4])])

table4=Table4(lineArr[0],lineArr[2],lineArr[5],'0')

tables_result.append(table4)

Merchant_ids_test[lineArr[2]]=testData.append([float(lineArr[3]), int(lineArr[4])])

#训练数据集读取

all_data = open('./train_data/Date_all.csv')

for line in all_data.readlines():

lineArr = line.strip().split(',')

features.append([float(lineArr[3]), int(lineArr[4])])

labels.append(int(lineArr[7]))

Merchant_ids_train[lineArr[2]]=trainData.append([float(lineArr[3]),int(lineArr[4]),int(lineArr[7])])

# print Merchant_ids_train.keys()

# print Merchant_ids_test.keys()

features_key=[] #数据集特征集

labels_key=[] #数据集类标集

num_not_in=0

for i in range(0,len(features_t)):

key=Merchant_ids_t[i]

key_dir_name='./train_data/merchant_train_data/'+str(key)+'_noNull'+'.csv'

features_key=[]

labels_key=[]

if os.path.exists(key_dir_name)==True:

key_data = open(key_dir_name)

for line in key_data.readlines():

lineArr = line.strip().split(',')

features_key.append([float(lineArr[3]), int(lineArr[4])])

labels_key.append(int(lineArr[7]))

print len(features_key)

print len(labels_key)

if len(features_key)>1:

Bay=BayesClassifier()

Bay.train(features_key,labels_key)

label,maxProbability=Bay.classify(features_t[i])

print("maxProbability:"+str(maxProbability)+"==>"+"Classified:"+label)

tables_result[i].giveProbability(str(maxProbability))

items=[tables_result[i].User_id,tables_result[i].Coupon_id,tables_result[i].Date_received,tables_result[i].Probability]

dir_name='./result/table4_4'

savecsv(dir_name,items)

else:

num_not_in=num_not_in+1

print num_not_in

# for line in file: #一行行读数据文件

# line=line.strip()

# tempVec=line.split(',')

# labels.append(tempVec[len(tempVec)-1])

# tempVec2=[tempVec[i] for i in range(0,len(tempVec)-1)]

# features.append(tempVec2)

# print len(features)

# print len(labels)

# features_n = features[0:940000]

# labels_n = labels[0:940000]

# Bay=BayesClassifier()

# Bay.train(features_n,labels_n)

# # correct=0

# # for i in range(0,len(features_t)):

# # label=Bay.classify(features_t[i])

# # print("Original:"+str(labels_t[i])+"==>"+"Classified:"+label)

# # if str(label)==str(labels_t[i]):

# # correct+=1

# # print correct

# # print len(features_t)

# # Accuracy=correct/len(features_t)

# # print "Accuracy:",Accuracy #正确率

# for i in range(0,len(features_t)):

# label,maxProbability=Bay.classify(features_t[i])

# print("maxProbability:"+str(maxProbability)+"==>"+"Classified:"+label)

# tables_result[i].giveProbability(str(maxProbability))

# items=[tables_result[i].User_id,tables_result[i].Coupon_id,tables_result[i].Date_received,tables_result[i].Probability]

# dir_name='./result/table4_2'

## step 1: load data

print "step 1: load data..."

dataSet = []

labels = []

# label_1 = open('./train_data/Date_label_1.csv')

# label_0 = open('./train_data/Date_label_0.csv')

# null_label_0 = open('./train_data/Date_null_label_0.csv')

# for line in label_1.readlines():

# lineArr = line.strip().split(',')

# dataSet.append([float(lineArr[3]), int(lineArr[4])])

# labels.append(int(lineArr[7]))

# for line in label_0.readlines():

# lineArr = line.strip().split(',')

# dataSet.append([float(lineArr[3]), int(lineArr[4])])

# labels.append(int('-1'))

# for line in null_label_0.readlines():

# lineArr = line.strip().split(',')

# dataSet.append([float(lineArr[3]), int(lineArr[4])])

# labels.append(int('-1'))

all_data = open('./train_data/Date_all.csv')

for line in all_data.readlines():

lineArr = line.strip().split(',')

dataSet.append([float(lineArr[3]), int(lineArr[4])])

labels.append(int(lineArr[7]))

# if lineArr[7]=='1':

# labels.append(int(lineArr[7]))

# else:

# labels.append(int('-1'))

print len(dataSet)

print len(labels)

dataSet_n = dataSet[0:40000]

labels_n = labels[0:40000]

# train_x = np.array(dataSet[0:60000])

# train_y = np.array(labels[0:60000])

label_num=-1

for data_x in dataSet_n:

label_num=label_num+1

if labels_n[label_num]== 1:

plt.plot(data_x[0], data_x[1], 'or')

elif labels_n[label_num]== 0:

plt.plot(data_x[0], data_x[1], 'ob')

plt.show()

#找数据规律可以发现，简单的：当距离固定，折扣率越大，核销率越大；

#一般同一个用户在不同时间领同一种优惠券，其用与不用是一样的

#

# test_x = dataSet[60000:60200]

# test_y = []

## step 2: training...

# print "step 2: training..."

# clf=SVC()

# clf.fit(train_x,train_y)

# print clf.predict([[0.95, 3]])

# for t_x in test_x:

# print clf.predict([t_x])

# test_y.append(clf.predict([t_x]))

# ## step 2: training...

# print "step 2: training..."

# C = 0.6

# toler = 0.001

# maxIter = 50

# svmClassifier = model.trainSVM(train_x, train_y, C, toler, maxIter, kernelOption = ('linear', 0))

# ## step 3: testing

# print "step 3: testing..."

# accuracy = model.testSVM(svmClassifier, test_x, test_y)

# ## step 4: show the result

# print "step 4: show the result..."

# print 'The classify accuracy is: %.3f%%' % (accuracy * 100)