# # Test Gaussian PDF
# print(calculate_probability(1.0, 1.0, 1.0))
# print(calculate_probability(2.0, 1.0, 1.0))
# print(calculate_probability(0.0, 1.0, 1.0))
# # Test calculating class probabilities
# summaries = summarize_by_class(dataset)
# row = dataset[0]
# probabilities = calculate_class_probabilities(summaries, row)
# print(probabilities)
# print(predict(summaries, row))
# Test Naive Bayes on Iris Dataset
seed(1)
dataset_base_path = os.path.join(os.path.dirname(os.getcwd()), 'datasets')
filename = 'iris.csv'
dataset = load_dataset(os.path.join(dataset_base_path, filename))
# Convert string to float
for i in range(len(dataset[0]) - 1):
str_column_to_float(dataset, i)
# convert class column to integers
str_column_to_int(dataset, len(dataset[0]) - 1)
# evaluate algorithm
n_folds = 5
scores = evaluate_algorithm(dataset, naive_bayes, 5, accuracy_metric)
print('Scores: {}'.format(scores))
print('Mean Accuracy: {:.3f}'.format(sum(scores) / float(len(scores))))
# Make a copy of test set to remove predictions while passing to the algorithm
test_set = deepcopy(test)
for row in test_set:
row[-1] = None
predicted = algorithm(train, test_set, *args)
actual = [row[-1] for row in test]
accuracy = metric(actual, predicted)
return accuracy
# zero rule algorithm for classification
def zero_rule_algorithm_classification(train, test):
output_values = [row[-1] for row in train]
prediction = max(set(output_values), key=output_values.count)
predicted = [prediction for _ in range(len(test))]
return predicted
if __name__ == '__main__':
# Test the train/test harness
seed(1)
dataset_base_path = os.path.join(os.path.dirname(os.getcwd()), 'datasets')
filename = 'pima-indians-diabetes.csv'
dataset = load_dataset(os.path.join(dataset_base_path, filename))
print('Loaded data file {0} with {1} rows and {2} columns'.format(
filename, len(dataset), len(dataset[0])))
for column in range(len(dataset[0])):
str_column_to_float(dataset, column)
accuracy = evaluate_algorithm(dataset, zero_rule_algorithm_classification,
0.6, accuracy_metric)
print('Accuracy {:.3f}'.format(accuracy))