from classification import train_classifier, test_classifier, cross_validation
from sklearn.ensemble import AdaBoostClassifier
if __name__ == '__main__':
classifier = AdaBoostClassifier()
train_classifier(classifier, select_features=True)
test_classifier(classifier, select_features=True)
print '------------------'
cross_validation(classifier, select_features=True)
from classification import train_classifier, test_classifier, cross_validation
from sklearn.tree import DecisionTreeClassifier
if __name__ == '__main__':
classifier = DecisionTreeClassifier()
train_classifier(classifier, select_features=False)
test_classifier(classifier, select_features=False)
print '------------------'
cross_validation(classifier, select_features=False)
print('Lexical features added\n')
# POS features
print('Preparing pos tag ratio vectors')
sarc_train_pos_ratios = flatten(pos_features.get_tag_ratio_vector(sarc_train))
sarc_test_pos_ratios = flatten(pos_features.get_tag_ratio_vector(sarc_test))
reg_train_pos_ratios = flatten(pos_features.get_tag_ratio_vector(reg_train))
reg_test_pos_ratios = flatten(pos_features.get_tag_ratio_vector(reg_test))
print('Parts of speech features added')
# Classification
print()
print('Preparing feature vectors of individual dataset')
sarc_train_features = classification.get_feature_vector([sarc_train_sent, sarc_train_rating]
+sarc_train_punctuations+sarc_train_pos_ratios)
sarc_test_features = classification.get_feature_vector([sarc_test_sent, sarc_test_rating]
+sarc_test_punctuations+sarc_test_pos_ratios)
reg_train_features = classification.get_feature_vector([reg_train_sent, reg_train_rating]
+reg_train_punctuations+reg_train_pos_ratios)
reg_test_features = classification.get_feature_vector([reg_test_sent, reg_test_rating]
+reg_test_punctuations+reg_test_pos_ratios)
print('Training and testing classifier')
classifier = classification.get_classifier(sarc_train_features, reg_train_features)
classification.test_classifier(classifier, sarc_test_features, reg_test_features)
# End of code
print('The end!')
review = {'rating': 1.0, 'review': 'I recently purchased this speaker and I want to tell you that it is totally worth the 1000 bucks!!! The voice is so clear and lound that I can hear voices even from other galaxies!'}
