linear_regression.py

# -*- coding: utf-8 -*-
# linear regression.

# I have completed the major framework, you can fill the space 
# to make it work.

__author__ = "Zifeng Wang"
__email__ = "wangzf18@mails.tsinghua.edu.cn"
__date__ = "20190920"

import numpy as np
import pdb
import os

np.random.seed(2019)


class linear_regression:

    def __init__(self):
        self.mse_func = lambda x, y: 1 / x.shape[0] * np.sum((x - y) ** 2)
        pass

    def train(self, x_train, y_train):
        """Receive the input training data, then learn the model.
        Inputs:
        x_train: np.array, shape (num_samples, num_features)
        y_train: np.array, shape (num_samples, )

        Outputs：
        None
        """

        self.learning_rate = 0.00001
        self.w = np.random.rand(10)
        iteration = 10000

        for i in range(iteration):
            pred = self.predict(x_train)

            """
            Please Fill Your Code Here.
            """
            self.w = self.w + self.learning_rate * (y_train[i] - x_train[i].dot(self.w)) * x_train[i]

            if i % 100 == 0:
                print("Iteration {}/{}, MSE loss {:.4f}".format(i + 1, iteration, self.mse_func(pred, y_train)))

        return

    def predict(self, x_test):
        """Do prediction via the learned model.
        Inputs:
        x_test: np.array, shape (num_samples, num_features)

        Outputs:
        pred: np.array, shape (num_samples, )
        """

        """
        Please Fill Your Code Here.
        """
        pred = x_test.dot(self.w)
        return pred


if __name__ == '__main__':

    # "load" your data
    real_weights = np.random.rand(10)
    x_train = np.random.randn(10000, 10) * 10
    y_train = x_train.dot(real_weights) + np.random.randn(10000)

    # train and test your model
    linear_regressor = linear_regression()
    linear_regressor.train(x_train, y_train)

    weights_error = np.sum((real_weights - linear_regressor.w) ** 2)
    print('weights error: ', weights_error)
    if weights_error < 1e-4:
        print("Congratulations! Your linear regressor works!")
    else:
        print("Check your code! Sth went wrong.")