def train_RF():
start_time = time.time()
cutoffLine('*')
print 'Use RF model to train %d models'%TRAIN_SET_FILES
for i in range(1, 1 + 1):
#for i in range(1, TRAIN_SET_FILES + 1):
cutoffLine('-')
print 'model %d'%i
t_file = file(TRAIN_SET_DIR + '/%d.csv'%i, 'r')
t_reader = csv.reader(t_file)
X = []
y = []
for line in t_reader:
line = map(int, line)
X.append(line[2:-1])
y.append(line[-1])
model = RF(X, y)
P ,R ,F = evaluate_model(model, i)
predict(model, i)
models.append(model)
t_file.close()
cutoffLine('*')
end_time = time.time()
duration = timekeeper(start_time, end_time)
print 'I takes %s to train , evaluate model and generate result'% duration
def sampling():
cutoffLine('*')
print 'Sampling using EasyEnsemble method'
start_time = time.time()
TRAIN_SET = 'training_set'
if not os.path.exists(TRAIN_SET): os.mkdir(TRAIN_SET)
propotion = 10
negative_size = POSITIVE * propotion
r_file = file(PRE_DIR + '/negative_set.csv', 'r')
reader = csv.reader(r_file)
positive_set = readCSV(PRE_DIR + '/positive_set.csv', int)
negative_set = []
set_count = 0
for line in reader:
progressBar(reader.line_num, NEGATIVE)
line = map(int, line)
if line[-1] == 1: positive_set.append(line)
if line[-1] == 0: negative_set.append(line)
if len(negative_set) == negative_size or reader.line_num == NEGATIVE:
set_count += 1
training_set = positive_set + negative_set
random.shuffle(training_set)
file_name = TRAIN_SET + '/' + '%d.csv'%set_count
writeCSV(training_set, file_name)
negative_set = []
r_file.close()
end_time = time.time()
duration = timekeeper(start_time, end_time)
cutoffLine('*')
print 'It takes %s to sampling' % duration
def generate_training_set():
start_time = time.time()
## load the information of data set
line_count = {}
rfile = file(PRE_DIR + '/stat.csv','r')
reader = csv.reader(rfile)
for line in reader:
line_count[line[0]] = int(line[1])
rfile.close()
cutoffLine('*')
print 'Generate training set'
for i in range(1,FILES + 1):
cutoffLine('-')
if i == FILES:
file_name = 'for_prediction.csv'
print 'Extract feature from %s'%file_name
extract_feature(file_name, line_count[file_name], i)
elif i == FILES - 1:
file_name = 'test.csv'
print 'Extract feature from %s'%file_name
result_name = 'result_%s'%file_name
extract_feature(file_name, line_count[file_name], i, result_name)
else:
file_name = '%d.csv' % i
print 'Extract feature from %s and tag it'%file_name
result_name = 'result_%d.csv' % i
extract_feature(file_name, line_count[file_name], i, result_name)
end_time = time.time()
duration = timekeeper(start_time, end_time)
cutoffLine('*')
print 'It takes %s to generate training set' % duration
def splitData():
cutoffLine('*')
print 'Start split data with window %d' % WINDOW
start_time = time.time()
stat_file = file(PRE_DIR + '/stat.csv', 'w')
stat_writer = csv.writer(stat_file)
for i in range(1, FILES + 1):
cutoffLine('-')
print 'Split dataset %d/%d: ' % (i, FILES)
rfile = file(DATA_SET, 'r')
reader = csv.reader(rfile)
j = i + WINDOW
if j != TOTAL_DAY + 1:
if j == TOTAL_DAY:
train_file_name = 'test.csv'
result_file_name = 'result_test.csv'
else:
train_file_name = '%d.csv' % i
result_file_name = '%s_%d.csv' % ('result', i)
train_file = file(PRE_DIR + '/' + train_file_name, 'w')
result_file = file(PRE_DIR + '/' + result_file_name, 'w')
train_writer = csv.writer(train_file)
result_writer = csv.writer(result_file)
train_count = 0
result_count = 0
for line in reader:
progressBar(reader.line_num, DATASET_SIZE)
if int(line[5]) >= i and int(line[5]) < j:
train_writer.writerow(line)
train_count += 1
if int(line[5]) == j and int(line[2]) == 4:
result_writer.writerow([line[0], line[1]])
result_count += 1
stat_writer.writerow([train_file_name, train_count])
stat_writer.writerow([result_file_name, result_count])
train_file.close()
result_file.close()
else:
forpredict_file_name = 'for_prediction.csv'
train_file = file(PRE_DIR + '/' + forpredict_file_name, 'w')
train_writer = csv.writer(train_file)
train_count = 0
for line in reader:
progressBar(reader.line_num, DATASET_SIZE)
if int(line[5]) >= i and int(line[5]) < j:
train_writer.writerow(line)
train_count += 1
stat_writer.writerow([forpredict_file_name, train_count])
train_file.close()
rfile.close()
stat_file.close()
end_time = time.time()
duration = timekeeper(start_time, end_time)
cutoffLine('-')
print 'It takes ' + duration + ' to split dataset.'
cutoffLine('*')
def splitData():
cutoffLine('*')
print 'Start split data with window %d' % WINDOW
start_time = time.time()
stat_file = file(PRE_DIR + '/stat.csv','w')
stat_writer = csv.writer(stat_file)
for i in range(1,FILES+1):
cutoffLine('-')
print 'Split dataset %d/%d: ' % (i, FILES)
rfile = file(DATA_SET,'r')
reader = csv.reader(rfile)
j = i + WINDOW
if j != TOTAL_DAY + 1:
if j == TOTAL_DAY:
train_file_name = 'test.csv'
result_file_name = 'result_test.csv'
else:
train_file_name = '%d.csv'%i
result_file_name = '%s_%d.csv'%('result',i)
train_file = file(PRE_DIR + '/' + train_file_name,'w')
result_file = file(PRE_DIR + '/' + result_file_name,'w')
train_writer = csv.writer(train_file)
result_writer = csv.writer(result_file)
train_count = 0
result_count = 0
for line in reader:
progressBar(reader.line_num, DATASET_SIZE)
if int(line[5]) >= i and int(line[5]) < j:
train_writer.writerow(line)
train_count += 1
if int(line[5]) == j and int(line[2]) == 4:
result_writer.writerow([line[0],line[1]])
result_count += 1
stat_writer.writerow([train_file_name, train_count])
stat_writer.writerow([result_file_name, result_count])
train_file.close()
result_file.close()
else:
forpredict_file_name = 'for_prediction.csv'
train_file = file(PRE_DIR + '/' + forpredict_file_name,'w')
train_writer = csv.writer(train_file)
train_count = 0
for line in reader:
progressBar(reader.line_num,DATASET_SIZE)
if int(line[5]) >= i and int(line[5]) < j:
train_writer.writerow(line)
train_count += 1
stat_writer.writerow([forpredict_file_name, train_count])
train_file.close()
rfile.close()
stat_file.close()
end_time = time.time()
duration = timekeeper(start_time,end_time)
cutoffLine('-')
print 'It takes ' + duration + ' to split dataset.'
cutoffLine('*')
def merge_training_set():
cutoffLine('*')
print 'Merge training set'
start_time = time.time()
positive_count = 0
negative_count = 0
total_count = 0
total_file = file(PRE_DIR + '/' + 'train_set.csv', 'w')
pos_file = file(PRE_DIR + '/' + 'positive_set.csv', 'w')
neg_file = file(PRE_DIR + '/' + 'negative_set.csv', 'w')
total_writer = csv.writer(total_file)
pos_writer = csv.writer(pos_file)
neg_writer = csv.writer(neg_file)
for i in range(1, FILES-1):
cutoffLine('-')
print 'load train set %d' % i
r_file = file(PRE_DIR + '/' + 'set_%d.csv' % i)
reader = csv.reader(r_file)
for line in reader:
doneCount(reader.line_num)
line = map(int, line)
if line[-1] == 1:
positive_count += 1
pos_writer.writerow(line)
if line[-1] == 0:
negative_count += 1
neg_writer.writerow(line)
total_count += 1
total_writer.writerow(line)
r_file.close()
total_file.close()
pos_file.close()
neg_file.close()
cutoffLine('-')
# 44114
print 'Positive Example: %d' % positive_count
# 59373295
print 'Negative Example: %d' % (total_count - positive_count)
# 59417409
print 'Total Example: %d' % total_count
# 一致性判断
print 'Is right? %s'%('Yes' if positive_count + negative_count == total_count else 'No')
end_time = time.time()
duration = timekeeper(start_time, end_time)
cutoffLine('*')
print 'It takes %s to merge training set and backup negative and positive set' % duration
def merge_training_set():
cutoffLine('*')
print 'Merge training set'
start_time = time.time()
positive_count = 0
negative_count = 0
total_count = 0
total_file = file(PRE_DIR + '/' + 'train_set.csv', 'w')
pos_file = file(PRE_DIR + '/' + 'positive_set.csv', 'w')
neg_file = file(PRE_DIR + '/' + 'negative_set.csv', 'w')
total_writer = csv.writer(total_file)
pos_writer = csv.writer(pos_file)
neg_writer = csv.writer(neg_file)
for i in range(1, FILES - 1):
cutoffLine('-')
print 'load train set %d' % i
r_file = file(PRE_DIR + '/' + 'set_%d.csv' % i)
reader = csv.reader(r_file)
for line in reader:
doneCount(reader.line_num)
line = map(int, line)
if line[-1] == 1:
positive_count += 1
pos_writer.writerow(line)
if line[-1] == 0:
negative_count += 1
neg_writer.writerow(line)
total_count += 1
total_writer.writerow(line)
r_file.close()
total_file.close()
pos_file.close()
neg_file.close()
cutoffLine('-')
print 'Positive Example: %d' % positive_count
print 'Negative Example: %d' % (total_count - positive_count)
print 'Total Example: %d' % total_count
print 'Is right? %s' % ('Yes' if positive_count +
negative_count == total_count else 'No')
end_time = time.time()
duration = timekeeper(start_time, end_time)
cutoffLine('*')
print 'It takes %s to merge training set and backup negative and positive set' % duration
def train(window, proportion, algo, confidence):
start_time = time.time()
cutoffLine('*')
print '%s model training with sample proportion 1:%d...' %(algo, proportion)
t_file = file('data/training_set_%d_%d.csv' % (window, proportion), 'r')
t_reader = csv.reader(t_file)
X = []
y = []
for line in t_reader:
doneCount(t_reader.line_num)
line = map(int, line)
X.append(line[3:-1])
y.append(line[-1])
model_name = 'data/model/%s_%d_%d.model'%(algo, window, proportion)
if os.path.exists(model_name): model = joblib.load(model_name)
else:
if algo == 'lr': model = LR(X, y)
if algo == 'rf': model = RF(X, y)
if algo == 'svm': model = SVM(X, y)
joblib.dump(model, model_name)
cutoffLine('-')
print model.classes_
item_subset = loadItemSubset()
record_file = open('data/model_evaluate_record.txt','a')
P, R, F= evaluate_model(algo, window, model, item_subset, confidence)
predict_set_size = predict(window, model, item_subset, proportion, algo, confidence)
record_file.write('window %d '%window + algo+' %d'%proportion + ' %.2f\n'%confidence)
record_file.write('\tP: %f\n'%P)
record_file.write('\tR: %f\n'%R)
record_file.write('\tF1: %f\n'%F)
record_file.write('Predict Set Size: %d\n'%predict_set_size)
record_file.write('-'*30+'\n')
record_file.close()
t_file.close()
cutoffLine('*')
end_time = time.time()
duration = timekeeper(start_time, end_time)
print 'I takes %s to train , evaluate model and generate result' % duration
def generate_training_set(window):
start_time = time.time()
global PRE_DIR, FILES
PRE_DIR = 'splited_data_%d' % window
FILES = TOTAL_DAY - window + 1
## load the information of data set
line_count = {}
rfile = file(PRE_DIR + '/stat.csv', 'r')
reader = csv.reader(rfile)
for line in reader:
line_count[line[0]] = int(line[1])
rfile.close()
cutoffLine('*')
print 'Generate training set with window %d' % window
for i in range(1, FILES + 1):
cutoffLine('-')
if i == FILES:
file_name = 'for_prediction.csv'
print 'Extract feature from %s' % file_name
extract_feature(window, i + window, file_name,
line_count[file_name], i)
elif i == FILES - 1:
file_name = 'test.csv'
print 'Extract feature from %s' % file_name
result_name = 'result_%s' % file_name
extract_feature(window, i + window, file_name,
line_count[file_name], i, result_name)
else:
file_name = '%d.csv' % i
print 'Extract feature from %s and tag it' % file_name
result_name = 'result_%d.csv' % i
extract_feature(window, i + window, file_name,
line_count[file_name], i, result_name)
end_time = time.time()
duration = timekeeper(start_time, end_time)
cutoffLine('*')
print 'It takes %s to generate training set' % duration
def train_LR():
start_time = time.time()
cutoffLine('*')
print 'LR model training...'
cutoffLine('-')
t_file = file('data/training_set_10.csv', 'r')
t_reader = csv.reader(t_file)
X = []
y = []
for line in t_reader:
line = map(int, line)
X.append(line[2:-1])
y.append(line[-1])
model = logRes(X,y)
item_subset = loadItemSubset()
evaluate_model(model, item_subset)
predict(model, item_subset)
t_file.close()
cutoffLine('*')
end_time = time.time()
duration = timekeeper(start_time, end_time)
print 'I takes %s to train , evaluate model and generate result' % duration
def sampling(proportion):
cutoffLine('*')
start_time = time.time()
print 'sampling with propotion %d...'%proportion
negative_needed = POSITIVE * proportion
sample_times = 10
mod = NEGATIVE / sample_times
negative_eachtime = negative_needed / sample_times
training_set = readCSV(PRE_DIR + '/positive_set.csv', int)
## sampling negative example
rfile = file(PRE_DIR + '/' + 'negative_set.csv', 'r')
reader = csv.reader(rfile)
negative_tmp = []
for line in reader:
progressBar(reader.line_num, NEGATIVE)
negative_tmp.append(map(int, line))
if reader.line_num % mod == 0:
random.shuffle(negative_tmp)
training_set = training_set + negative_tmp[0:negative_eachtime]
negative_tmp = []
rfile.close()
wfile = file('data/training_set_%d.csv'%proportion, 'w')
writer = csv.writer(wfile)
random.shuffle(training_set)
writer.writerows(training_set)
wfile.close()
cutoffLine('-')
print "Real proportion: %f" %((len(training_set)-POSITIVE) / float(POSITIVE))
cutoffLine('*')
end_time = time.time()
duration = timekeeper(start_time, end_time)
print 'It takes %s to sampling with proportion %d'%(duration, proportion)
def sampling(window, proportion):
cutoffLine('*')
start_time = time.time()
print 'sampling with propotion %d...' % proportion
exec('negative_needed = POSITIVE_%d * propotion' % window)
sample_times = 20
exec('mod = NEGATIVE_%d / sample_times' % window)
exec('negative_eachtime = negative_needed / sample_times')
training_set = readCSV(PRE_DIR + '/positive_set.csv', int)
## sampling negative example
rfile = file(PRE_DIR + '/' + 'negative_set.csv', 'r')
reader = csv.reader(rfile)
negative_tmp = []
for line in reader:
exec('progressBar(reader.line_num, NEGATIVE_%d)' % window)
negative_tmp.append(map(int, line))
if reader.line_num % mod == 0:
random.shuffle(negative_tmp)
training_set.extend(negative_tmp[0:negative_eachtime])
negative_tmp = []
rfile.close()
wfile = file('data/training_set_%d_%d.csv' % (window, propotion), 'w')
writer = csv.writer(wfile)
random.shuffle(training_set)
writer.writerows(training_set)
wfile.close()
cutoffLine('-')
exec('real_proportion = (len(training_set)- POSITIVE_%d) / float(POSITIVE_%d)'%(window, window))
print "Real proportion: %f" % real_proportion
cutoffLine('*')
end_time = time.time()
duration = timekeeper(start_time, end_time)
print 'It takes %s to sampling with proportion %d'%(duration, proportion)
if int(line[5]) == j and int(line[2]) == 4:
result_writer.writerow([line[0],line[1]])
result_count += 1
stat_writer.writerow([train_file_name, train_count])
stat_writer.writerow([result_file_name, result_count])
train_file.close()
result_file.close()
else:
forpredict_file_name = 'for_prediction.csv'
train_file = file(PRE_DIR + '/' + forpredict_file_name,'w')
train_writer = csv.writer(train_file)
train_count = 0
for line in reader:
progressBar(reader.line_num,DATASET_SIZE)
if int(line[5]) >= i and int(line[5]) < j:
train_writer.writerow(line)
train_count += 1
stat_writer.writerow([forpredict_file_name, train_count])
train_file.close()
rfile.close()
if __name__ == '__main__':
print 'Start split data'
cutoffLine('*')
start_time = time.time()
splitData()
end_time = time.time()
duration = timekeeper(start_time,end_time)
cutoffLine('*')
print 'It takes ' + duration + ' to split dataset.'