def _evaluate_pos_tagger():
words, real_tags = _get_data()
tagger = POSTagger()
_w, pred_tags = _split_tagged_words(tagger.tag(words))
for i in range(100):
print(words[i], real_tags[i], pred_tags[i])
return accuracy(real_tags, pred_tags)
def test_bool_66():
y_true = [True, False, True]
y_pred = [True, False, False]
assert accuracy(y_true, y_pred) == 0.6667
def test_str():
y_true = ['a', 'b', 'c']
y_pred = ['d', 'b', 'a']
assert accuracy(y_true, y_pred) == 0.3333
def test_list():
y_true = [['a', 'b'], ['c']]
y_pred = [['a', 'b'], ['d']]
assert accuracy(y_true, y_pred) == 0.5
def test_bool_100():
y_true = [True, False, True]
y_pred = [True, False, True]
assert accuracy(y_true, y_pred) == 1.0
def test_float():
y_true = [10.4, 4.7, 3.0, 5.02]
y_pred = [10.40, 4.74, 3, 5.02]
assert accuracy(y_true, y_pred) == 0.75
def test_int():
y_true = [10, 4, 3, 5]
y_pred = [3, 4, 7, 5]
assert accuracy(y_true, y_pred) == 0.5
def _evaluate_classifier(name):
x_train, y_train, x_test, y_test = _get_data()
classifier = name()
classifier.train(x_train, y_train)
y_pred = classifier.classify(x_test)
return accuracy(y_test, y_pred)
def evaluate_classifier(y_true, y_prediction):
print('* Accuracy: {:.2%}'.format(accuracy(y_true, y_prediction)))
