def main():
file=open("Bupa.txt",'r')
features=[] #数据集特征集
labels=[] #数据集类标集
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)
Bay=BayesClassifier()
Bay.train(features,labels)
correct=0
for i in range(0,len(features)):
label=Bay.classify(features[i])
print("Original:"+str(labels[i])+"==>"+"Classified:"+label)
if str(label)==str(labels[i]):
correct+=1
print("Accuracy:",correct/len(features)) #正确率
def testBayes():
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
