#pragma once

#ifndef ML_LINEAEEEGRESSION_H

#define ML_LINEARREGRESSION_H

class LinearRegression {

public:

    /*特征*/

    double *x;

    /*预测值*/

    double *y;

    /*样本数量*/

    int m;

    /*系数*/

    double *theta;

    /*创建实例*/

    LinearRegression(double x[], double y[], int m);

    /*训练 */

    void train(double alpha, int iterations);

    /*预测*/

    double predict(double x);

private:

    /*计算损失模型*/

    static double compute_cost(double x[], double y[], double theta[], int m);

    /*计算单个预测值*/

    static double h(double x, double theta[]);

    /*预测*/

    static double *calculate_predictions(double x[], double theta[], int m);

    /*梯度下降*/

    static double *gradient_descent(double x[], double y[], double alpha, int iter, double *j, int m);

};

#endif // !ML_LINEAEEEGRESSION_H

#include "iostream"

#include "linearRegression.h"

#include "Utils.h"

using namespace std;

/*初始化*/

LinearRegression::LinearRegression(double x[], double y[], int m)

{

    this->x = x;

    this->y = y;

    this->m = m;

}

/*

alpha:learn rate

iterations:iterators

*/

void LinearRegression::train(double alpha, int iterations)

{

    double *J = new double[iterations];

    this->theta = gradient_descent(x, y, alpha, iterations, J, m);

    cout << "J=";

    for (int i = ; i < iterations; ++i)

    {

        cout << J[i] << " " << endl;;

    }

    cout << "\n" << "Theta: " << theta[] << " " << theta[] << endl;

}

/*预测*/

double LinearRegression::predict(double x)

{

    cout << "y':" << h(x, theta) << endl;

    return h(x, theta);

}

/*计算损失模型*/

double LinearRegression::compute_cost(double x[], double y[], double theta[], int m)

{

    double *predictions = calculate_predictions(x, theta, m);

    double *diff = Utils::array_diff(predictions, y, m);

    double *sq_errors = Utils::array_pow(diff, m, );

    return (1.0 / ( * m))*Utils::array_sum(sq_errors, m);

}

/*计算单个预测值*/

double LinearRegression::h(double x, double theta[])

{

    return theta[] + theta[] * x;

}

/*预测*/

double *LinearRegression::calculate_predictions(double x[], double theta[], int m)

{

    double *predictions = new double[m];

    for (int i = ; i < m; i++)

    {

        predictions[i] = h(x[i], theta);

    }

    return predictions;

}

/*梯度下降*/

double *LinearRegression::gradient_descent(double x[], double y[], double alpha, int iter, double *J, int m)

{

    double *theta = new double[];

    theta[] = ;

    theta[] = ;

    for (int i = ; i < iter; i++)

    {

        double *predictions = calculate_predictions(x, theta, m);

        double *diff = Utils::array_diff(predictions, y, m);

        double *error_x1 = diff;

        double *error_x2 = Utils::array_multiplication(diff, x, m);

        /*这里可以设定J损失函数的阈值,也可以设定梯度变化量的阈值*/

        theta[] = theta[] - alpha*(1.0 / m) * Utils::array_sum(error_x1, m);

        theta[] = theta[] - alpha*(1.0 / m)*Utils::array_sum(error_x2, m);

        J[i] = compute_cost(x, y, theta, m);

    }

    return theta;

}

#include "iostream"

#include "linearRegression.h"

using namespace std;

int main()

{

    double x[] = {,,,,};

    double y[] = {,,,,};

    LinearRegression test(x,y,);

    test.train(0.1, );

    test.predict();

    system("pause");

    return ;

}