from perceptron import Perceptron
# defining a
a = [
-1, -1, 1, -1, 1, -1, 1, -1, 1, -1, -1, 1, 1, 1, -1, 1, -1, -1, -1, 1, 1,
-1, -1, -1, 1
]
# defining A inverted
ai = [
1, -1, -1, -1, 1, 1, -1, -1, -1, 1, -1, 1, 1, 1, -1, -1, 1, -1, 1, -1, -1,
-1, 1, -1, -1
]
if __name__ == '__main__':
neural_network = Perceptron(5 * 5)
train_cases = [neural_network.generate_data(a) for _ in range(10)
] + [neural_network.generate_data(ai) for _ in range(10)]
train_labels = [-1 for _ in range(10)] + [1 for _ in range(10)]
test_cases = [neural_network.generate_data(a) for _ in range(5)
] + [neural_network.generate_data(ai) for _ in range(5)]
test_labels = [-1 for _ in range(5)] + [1 for _ in range(5)]
neural_network.train(train_cases, train_labels, 100, 0.01)
outs = [neural_network.recognize(x) for x in test_cases]
accuracy = sum(
[1 if y == label else 0 for y, label in zip(outs, test_labels)])
neural_network.write_text(train_cases, 'train_data.txt')
neural_network.write_text(train_labels, 'train_labels.txt')
neural_network.write_text(test_cases, 'test_data.txt')
neural_network.write_text(test_labels, 'test_labels.txt')
print('accuracy:', accuracy / len(outs))
