Gradient descent in python

gradient_descent.py

import numpy as np

X = [1, 2, 4, 3, 5]
y = [1, 3, 3, 2, 5]
alpha = 0.01
B_0 = 0
B_1 = 0
EPOCHS = 10

def predict(X_i, B_0, B_1):
	y_i = B_0 + B_1 * X_i
	return y_i

def calculate_error(X_i, y_i, B_0, B_1):
	return predict(X_i, B_0, B_1) - y_i

def calculate_B_0(X_i, y_i, B_0, B_1, alpha):
	error = calculate_error(X_i, y_i, B_0, B_1)
	B_n = B_0 - alpha * error 
	return B_n

def calculate_B_1(X_i, y_i, B_0, B_1, alpha):
	error = calculate_error(X_i, y_i, B_0, B_1)
	B_n = B_1 - alpha * error * X_i
	return B_n

for i in range(EPOCHS):
	for i, X_i in enumerate(X):
		y_i = y[i]
		B_0 = calculate_B_0(X_i, y_i, B_0, B_1, alpha)
		B_1 = calculate_B_1(X_i, y_i, B_0, B_1, alpha)
		print(B_0, B_1)


# B_0 = 0.2420425534906209
# B_1 = 0.8182420919125936
def f(X, y):
	y = B_0 + B_1 * X
	return y

y_pred = []
for i, X_i in enumerate(X):
	pred = f(X_i, y[i])
	y_pred.append(pred)
	print(pred)

def RMSE(X, y):
	return np.sqrt(np.mean(np.square(X - y)))

print(RMSE(np.array(X), np.array(y_pred)))