def test_whenNoPredictionIsCorrectAccuracyIs0(self): rewardClasses = [0,1] confusionMatrix = ConfusionMatrix(rewardClasses) realReward = 0 predictedReward = 1 confusionMatrix.addObservation(realReward, predictedReward) self.assertEqual(0, confusionMatrix.accuracy())
def test_whenAllPredictionsAreCorrectAccuracyIs100(self): rewardClasses = [0,1] confusionMatrix = ConfusionMatrix(rewardClasses) realReward = 0 predictedReward = 0 confusionMatrix.addObservation(realReward, predictedReward) self.assertEqual(100, confusionMatrix.accuracy())
def test_whenObservationIsAddedCountIncreasesByOne(self):
rewardClasses = [0,1]
confusionMatrix = ConfusionMatrix(rewardClasses)
realReward = 0
predictedReward = 0
confusionMatrix.addObservation(realReward, predictedReward)
self.assertEqual(1, confusionMatrix.confusionMatrix[realReward][predictedReward])
def test_forest_list(forest_list, cv, database):
classes = db.get_target_attribute_values()
f1measure = 0
for i, forest in enumerate(forest_list):
test_set = cv.folds[i]
(real_labels, pred_labels) = forest.test(test_set, debug)
confusion_matrix = ConfusionMatrix.create_confusion_matrix(
real_labels, pred_labels, classes)
f1measure += Metrics.f1measure(confusion_matrix)
if debug:
print('Testing the forest with the dataset:')
print(test_set)
print('Test Results:')
print('%d: ' % (i + 1))
print(confusion_matrix)
f1measure /= len(forest_list)
if debug:
print('Mean F1-Measure: ', f1measure)
return f1measure
def test_whenConfusionMatrixAreMergedCountsAreAdded(self):
rewardClasses = [0,1]
confusionMatrix1 = ConfusionMatrix(rewardClasses)
confusionMatrix2 = ConfusionMatrix(rewardClasses)
realReward = 0
predictedReward = 0
confusionMatrix2.addObservation(realReward, predictedReward)
confusionMatrix1.merge(confusionMatrix2)
self.assertEqual(1, confusionMatrix1.confusionMatrix[realReward][predictedReward])
real_classes = [
'c1', 'c1', 'c1', 'c1', 'c1', 'c1', 'c1', 'c1', 'c1', 'c1', 'c1', 'c1',
'c1', 'c1', 'c1', 'c2', 'c2', 'c2', 'c2', 'c2', 'c3', 'c3', 'c3', 'c3',
'c3', 'c3', 'c3', 'c3', 'c3', 'c3'
]
pred_classes_array = [
'c1', 'c1', 'c1', 'c1', 'c1', 'c1', 'c1', 'c1', 'c1', 'c1', 'c1', 'c2',
'c3', 'c3', 'c3', 'c2', 'c2', 'c2', 'c2', 'c1', 'c3', 'c3', 'c3', 'c3',
'c3', 'c3', 'c1', 'c1', 'c2', 'c2'
]
classes = ['c1', 'c2', 'c3']
confusion_matrix = ConfusionMatrix.create_confusion_matrix(
real_classes, pred_classes_array, classes)
sklean_confusion_matrix = metrics.confusion_matrix(real_classes,
pred_classes_array,
classes)
print(confusion_matrix)
print(sklean_confusion_matrix)
f1measure = f1measure(confusion_matrix)
sklearn_f1measure = metrics.precision_recall_fscore_support(
real_classes, pred_classes_array, average='macro')
print('f1measure\t', f1measure)
print('sk fb_score\t', sklearn_f1measure[2])
