# randomize
random.seed(i)
random.shuffle(dataset)
# split
train = dataset[:9*len(dataset)/10]
test = dataset[9*len(dataset)/10:]
# build classifiers
nbc = NaiveBayesClassifier(train)
dtc = DecisionTreeClassifier(train)
# save results
results_nbc.append(nbc.accuracy(test));
results_dtc.append(dtc.accuracy(test));
# output the mean of accuray
print('mean of accuracy:')
print('naive bayes', np.array(results_nbc).mean())
print('decision tree', np.array(results_dtc).mean())
# 2. use test_deals.txt for classification
nbc = NaiveBayesClassifier(dataset)
dtc = DecisionTreeClassifier(dataset)
print('naive bayes classification:')
for text in test_dat:
print(text, nbc.classify(text))
