LinearRegression.java

import java.io.File;
import java.io.FileNotFoundException;
import java.util.ArrayList;
import java.util.Scanner;

public class LinearRegression {
    public static void main(String[] args) {
       
        try{
            double[][] data = readPizzaData("pizza.txt");
            double[] X = data[0];
            double[] Y = data[1];

            LinearRegression model = new LinearRegression();

            int iterations = 20000;
            double lr = 0.001;

            double[] params = model.train(X, Y, iterations, lr);
            System.out.printf("\nw = %.10f, b = %.10f\n", params[0], params[1]);
            System.out.printf("Prediction: x = %d => y = %.2f%n", 20, model.predict(20, params[0], params[1]));
        } catch (FileNotFoundException e){
            System.out.println("File not found: " + e.getMessage());
        }
    }

    public static double [][] readPizzaData(String file) throws FileNotFoundException{
        Scanner sc = new Scanner(new File(file));

        if (sc.hasNextLine()) sc.nextLine();

        ArrayList<Double> Reservations = new ArrayList<>();
        ArrayList<Double> Pizzas = new ArrayList<>();

        while(sc.hasNext()){
            if(!sc.hasNextDouble()){
                sc.next();
                continue;
            }
            double r = sc.nextDouble();

            if(!sc.hasNextDouble()){
                break;
            }
            double p = sc.nextDouble();

            Reservations.add(r);
            Pizzas.add(p);
        }
        sc.close();

        double[] X = new double[Reservations.size()];
        double[] Y = new double[Pizzas.size()];

        for(int i = 0; i < Reservations.size(); i++){
            X[i] = Reservations.get(i);
            Y[i] = Pizzas.get(i);
        }
        return new double[][] {X, Y};
    }

    public double predict(double x, double w, double b){
        return x * w + b;
    }

    public double loss(double[] X, double[] Y, double w, double b){
        double sum = 0.0;
        int n = X.length;

        for (int i = 0; i < n; i++) {
            double err = predict(X[i], w, b) - Y[i];
            sum += err * err;
        }
        return sum / n;
    }

    public double[] gradient(double[] X, double[] Y, double w, double b){
        double sumWXErr = 0.0;
        double sumErr = 0.0;
        int n = X.length;

        for (int i = 0; i < n; i++) {
            double err = (X[i] * w + b) - Y[i];
            sumWXErr += X[i] * err;
            sumErr += err;
        }
        double wGrad = 2.0 * (sumWXErr / n);
        double bGrad = 2.0 * (sumErr / n);

        return new double[] {wGrad, bGrad};
    }

    public double[] train(double[] X, double[] Y, int iterations, double lr){
        double w = 0.0;
        double b = 0.0;

        for(int i = 0; i < iterations; i++){
            System.out.printf("Iteration %4d => Loss: %.10f\n", i, loss(X, Y, w, b));
            double[] grads = gradient(X, Y, w, b);
            w -= lr * grads[0];
            b -= lr * grads[1];
        }
        return new double[] {w,b};
    }
}