for doc in doclist:
print 'test doc: '+doc.encode('gbk')+'\n'
doc_catalog_pro = {}
wordlist = doclist[doc]
for catalog_try in catalog_p:
likely = catalog_p[catalog_try]
for word, value in features:
if word in wordlist:
likely += catalog_word_p[catalog_try][word]
doc_catalog_pro[catalog_try] = likely
nb_prediction[doc] = doc_catalog_pro
catalog_predicted = get_predicted_catalog(nb_prediction)
return catalog_predicted
if __name__ == '__main__':
ans = predict_nb(test_doc_to_word, FeatureFiltering.global_features)
print 'finished predicted by naive bayes'
true_positive, predicted_num, accuracy = KNNPredict.get_accuracy(ans)
file = open('E:\\TextClassificationData\\nb_content.txt', 'w')
file.write('true_positive, predicted_num, accuracy, %d %d %6lf \n' \
%(true_positive, predicted_num, accuracy))
for doc in ans:
file.write(doc.encode('gbk')+'\n')
file.write(' '+ans[doc].encode('gbk')+'\n')
file.close()
for doc in doclist:
print 'test doc: ' + doc.encode('gbk') + '\n'
doc_catalog_pro = {}
wordlist = doclist[doc]
for catalog_try in catalog_p:
likely = catalog_p[catalog_try]
for word, value in features:
if word in wordlist:
likely += catalog_word_p[catalog_try][word]
doc_catalog_pro[catalog_try] = likely
nb_prediction[doc] = doc_catalog_pro
catalog_predicted = get_predicted_catalog(nb_prediction)
return catalog_predicted
if __name__ == '__main__':
ans = predict_nb(test_doc_to_word, FeatureFiltering.global_features)
print 'finished predicted by naive bayes'
true_positive, predicted_num, accuracy = KNNPredict.get_accuracy(ans)
file = open('E:\\TextClassificationData\\nb_content.txt', 'w')
file.write('true_positive, predicted_num, accuracy, %d %d %6lf \n' \
%(true_positive, predicted_num, accuracy))
for doc in ans:
file.write(doc.encode('gbk') + '\n')
file.write(' ' + ans[doc].encode('gbk') + '\n')
file.close()
# coding: GBK
import KNNPredict
import Training
import configuration
if __name__ == '__main__':
testFileVectorFile = open('E:\\TextClassificationData\\test_content.txt',
'w')
testFileVector = None
print 'began to get testFileVector'
testFileVector = KNNPredict.getDocVector(KNNPredict.test_files_to_words,
Training.featureVector)
print 'finished getting testFileVector'
print 'began to predicting by knn method'
result, prediction = KNNPredict.KNN(KNNPredict.training_doc_vector,
testFileVector,
configuration.top_k_number)
true_positive, predicted, accuracy = KNNPredict.get_accuracy(prediction)
print 'finished predicting'
testFileVectorFile.write('%d %d accuracy:%.6lf' %
(true_positive, predicted, accuracy) + '\n')
for doc in result:
testFileVectorFile.write(
str(doc).encode('gbk') + ' ' + str(prediction[doc]).encode('gbk') +
'\n')
vector = result[doc]
for neighbor in vector:
testFileVectorFile.write(' ' + neighbor[0].encode('gbk') + '\n')
# coding: GBK
import KNNPredict
import Training
import configuration
if __name__ == '__main__':
testFileVectorFile = open('E:\\TextClassificationData\\test_content.txt', 'w')
testFileVector = None
print 'began to get testFileVector'
testFileVector = KNNPredict.getDocVector(KNNPredict.test_files_to_words, Training.featureVector)
print 'finished getting testFileVector'
print 'began to predicting by knn method'
result, prediction = KNNPredict.KNN(KNNPredict.training_doc_vector, testFileVector, configuration.top_k_number)
true_positive, predicted, accuracy = KNNPredict.get_accuracy(prediction)
print 'finished predicting'
testFileVectorFile.write('%d %d accuracy:%.6lf'%(true_positive, predicted, accuracy)+'\n')
for doc in result:
testFileVectorFile.write(str(doc).encode('gbk')+' '+str(prediction[doc]).encode('gbk')+'\n')
vector = result[doc]
for neighbor in vector:
testFileVectorFile.write(' '+neighbor[0].encode('gbk')+'\n')
testFileVectorFile.write(' '+str(neighbor[1])+'\n')
testFileVectorFile.close()
