def test_MissingPredictions(self):
qa = {"P1": "E", "P2": "N", "P3": "N"}
p = {"P1": "E", "P2": "N"}
with self.assertRaises(SystemExit) as context:
evaluator.calculate_accuracy(qa, p)
self.assertEqual(context.exception.code, evaluator.EXIT_STATUS_PREDICTION_MISSING)
def main():
print 'Accuracies'
print 'k\tBinary \tMulticlass'
for k in range(3, 7):
# 2 class classification
X_train, y_train = preprocessor.get_binary_train_features(k)
mlr_model = logisticRegression.perform_binary_logistic_regression(
X_train, y_train)
X_test, y_test = preprocessor.get_binary_test_features(k)
pred_y = mlr_model.predict(X_test)
binary_accuracy = evaluator.calculate_accuracy(y_test, pred_y)
binary_accuracy = round(binary_accuracy, 2)
#Classification with 3 classes
X_train, y_train = preprocessor.get_multiclass_train_features(k)
mlr_model = logisticRegression.perform_multiclass_logistic_regression(
X_train, y_train)
X_test, y_test = preprocessor.get_multiclass_test_features(k)
pred_y = mlr_model.predict(X_test)
multiclass_accuracy = evaluator.calculate_accuracy(y_test, pred_y)
multiclass_accuracy = round(multiclass_accuracy, 2)
print str(k) + '\t' + str(binary_accuracy) + ' \t' + str(
multiclass_accuracy)
def test_ExtraPredictions(self):
qa = {"Q1": "A", "Q2": "A", "Q3": "A"}
p = {"Q1": ["A"], "Q2": ["A"], "Q3": ["B"], "QExtra": ["X"]}
with self.assertRaises(SystemExit) as context:
evaluator.calculate_accuracy(qa, p)
self.assertEqual(context.exception.code,
evaluator.EXIT_STATUS_PREDICTIONS_EXTRA)
def main():
print 'Accuracies'
print 'k\tBinary \tMulticlass'
for k in range(3,7):
# 2 class classification
X_train, y_train = preprocessor.get_binary_train_features(k)
mlr_model = logisticRegression.perform_binary_logistic_regression(X_train, y_train)
X_test, y_test = preprocessor.get_binary_test_features(k)
pred_y = mlr_model.predict(X_test)
binary_accuracy = evaluator.calculate_accuracy(y_test, pred_y)
binary_accuracy = round(binary_accuracy, 2)
#Classification with 3 classes
X_train, y_train = preprocessor.get_multiclass_train_features(k)
mlr_model = logisticRegression.perform_multiclass_logistic_regression(X_train, y_train)
X_test, y_test = preprocessor.get_multiclass_test_features(k)
pred_y = mlr_model.predict(X_test)
multiclass_accuracy = evaluator.calculate_accuracy(y_test, pred_y)
multiclass_accuracy = round(multiclass_accuracy, 2)
print str(k) + '\t' + str(binary_accuracy) + ' \t' + str(multiclass_accuracy)
def test_EverythingCorrect(self):
qa = {"P1": "E", "P2": "N", "P3": "N"}
p = {"P1": "E", "P2": "N", "P3": "N"}
self.assertEqual(3.0 / 3.0, evaluator.calculate_accuracy(qa, p))
def test_MixedResults(self):
qa = {"P1": "E", "P2": "N", "P3": "N"}
p = {"P1": "E", "P2": "N", "P3": "E"}
self.assertEqual(2.0 / 3.0, evaluator.calculate_accuracy(qa, p))
def test_EverythingCorrect(self):
qa = {"Q1": "A", "Q2": "A", "Q3": "A"}
p = {"Q1": ["A"], "Q2": ["A"], "Q3": ["A"]}
self.assertEqual(3.0 / 3.0, evaluator.calculate_accuracy(qa, p))
def test_PartialGuess(self):
qa = {"Q1": "A", "Q2": "A", "Q3": "A"}
p = {"Q1": ["A", "B"], "Q2": ["B"], "Q3": ["B"]}
self.assertEqual(0.5 / 3, evaluator.calculate_accuracy(qa, p))
def test_MixedResults(self):
qa = {"Q1": "A", "Q2": "A", "Q3": "A"}
p = {"Q1": ["A"], "Q2": ["A"], "Q3": ["B"]}
self.assertEqual(2.0 / 3.0, evaluator.calculate_accuracy(qa, p))
