/** @file xor.c
  * @brief A test program for the Neural Network library to learn the XOR function.
  * The program trains the network on the XOR function and then tests its accuracy.
  * The model is a very simple one with only one hidden layer of 3 neurons.
  * The accuracy is then printed using the mean squared errror calculation.
  *
  *  @author Saleh Bubshait
  */

#include "../src/network.h"

int main(void) {
    int dimensions[] = {2, 3, 1};
    Network *network = network_create(3, dimensions);

    double inputs[4][2] = {{0, 0}, {0, 1}, {1, 0}, {1, 1}};
    double targets[4][1] = {{0}, {1}, {1}, {0}};

    int epochs = 15000;
    for (int i = 0; i < epochs; i++) {
        for (int j = 0; j < 4; j++) {
            Matrix *input = matrix_create(2, 1);
            matrix_initialise(input, (double[2][1]){inputs[j][0], inputs[j][1]});

            Matrix *target = matrix_create(1, 1);
            matrix_initialise(target, (double[1][1]){targets[j][0]});

            network_train(network, input, target, 0.1);

            matrix_free(input);
            matrix_free(target);
        }
    }

    double error = 0;
    for (int i = 0; i < 4; i++) {
        Matrix *input = matrix_create(2, 1);
        matrix_initialise(input, (double[2][1]){inputs[i][0], inputs[i][1]});

        network_predict(network, input);
        printf("Input: %d %d, Output: %f\n", (int)inputs[i][0], (int)inputs[i][1], network_output(network)[0]);
        error += (network_output(network)[0] - targets[i][0]) * (network_output(network)[0] - targets[i][0]);

        matrix_free(input);
    }

    printf("====================================\n");
    printf("Accuracy: %lf \n", 1 - error / 4); // Mean Squared Error

    network_free(network);

    return 0;
}