mvl.py

import numpy as np 
import pandas as pd 

df = pd.read_csv('m_dataset.txt', header=None, names=['size', 'age', 'price'])
df = (df - df.mean()) / df.std()

df.insert(0, 'Ones', 1)
cols = df.shape[1]

X = np.matrix(df.iloc[:, 0:cols-1].values)
y = np.matrix(df.iloc[:, cols-1:cols].values)

theta = np.matrix(np.array([0, 0, 0]))

def compute_cost(X, y, theta):
	hypothesis = np.power(np.matmul(X, theta.T) - y, 2)
	return np.sum(hypothesis) / 2 * len(X)


def gradient_descent(X, y, theta, alpha, iters):
	temp_theta = np.matrix(np.zeros(theta.shape))
	parameters = int(theta.ravel().shape[1])
	cost = np.zeros(iters)

	for i in range(iters):
		error = (X * theta.T) - y	

		for j in range(parameters):
			column = np.multiply(error, X[:, j])
			temp_theta[0, j] = theta[0, j] - ((alpha / len(X)) * np.sum(column))

		
		final_theta = temp_theta
		cost[i] = compute_cost(X, y, final_theta)

	return final_theta, cost


alpha = 0.001
iters = 1000

params, cost = gradient_descent(X, y, theta, alpha, iters)
print(params)