single_value_linear_regression.py

# -*- coding: utf-8 -*-
""" MIT License """

import matplotlib.pyplot as plt
import numpy as np


X = [10,9,8,7,6,5,4,3,2,1,0]
Y = [1,2,4,3,5,4,6,7,4,8,9]

m = len(X)
alpha = 0.01

def h(x, theta):
    return theta[0] + theta[1] * x

def J(theta):
    sum = 0
    for i in range(m):
        sum += (h(X[i], theta) - Y[i])**2
        
    return sum / m;    

def J_partial0(theta):
    sum = 0
    for i in range(m):
        sum += h(X[i], theta) - Y[i]
        
    return 2 * sum / m;
    
def J_partial1(theta):
    sum = 0
    for i in range(m):
        sum += (h(X[i], theta) - Y[i]) * X[i]
        
    return 2 * sum / m;

def gradient_descent(startTheta):
    theta = startTheta
    for i in range(5000):
        d0 = J_partial0(theta)
        d1 = J_partial1(theta)
        theta[0] -= alpha * d0
        theta[1] -= alpha * d1
    
    return theta


t = gradient_descent([0,0])

XX = np.linspace(0,10,1000)
YY = h(XX, t)
plt.plot(XX,YY)
plt.scatter(X,Y, color='r')
plt.xlabel('Opuszczone wykłady')
plt.ylabel('Punkty')
plt.show()