LinearRegessor.py

import numpy as np
class LinearRegressor():
    def __init__(self):
        self.θ0 = np.random.random((1,))[0]
        self.θ1 = np.random.random((1,))[0]
    
    def predict(self, x):
        return self.θ0 + self.θ1 * np.array(x)
    
    def compute_loss(self, y_preds, y_true):
        return (sum((y_preds - y_true) ** 2)) / len(y_preds)
    
    def fit(self, x, y, epochs, learning_rate=0.001, seed=42):
        np.random.seed(seed)
        θ0, θ1 = np.random.random((2,))
        for i in range(1, epochs+1):
            y_preds = self.predict(x)
            self.θ0 = θ0 - learning_rate * (2 / len(y)) * sum(y_preds - y)
            self.θ1 = θ1 - learning_rate * sum((y_preds - y) * x)
            print(f"Epoch {i}:\nLoss: {round(self.compute_loss(y_preds, y), 4)}")