线性回归算法Python小实现

demo:数据集

#!/usr/bin/env python3
# -*- coding: utf-8 -*-
"""
Created on Fri Mar  9 20:05:51 2018

@author: lisir
"""

import csv
import numpy as np
import matplotlib.pyplot as plt
import matplotlib as mpl
import pandas as pd
from sklearn.model_selection import train_test_split
from sklearn.linear_model import LinearRegression
from pprint import pprint

if __name__ == "__main__":
    csv_path = "/home/lisir/test/c_test/ml/tmp/10.Regression/Advertising.csv"
    # read data method one
    f = open(csv_path)
    x = []
    y = []
    for i, d in enumerate(f):
        if i == 0:
            continue
        d = d.strip()
        if not d:
            continue
        # split data by ,
        data = list(map(float, d.split(',')))
        x.append(data[1:3])
        y.append(data[-1])
    pprint(x)
    pprint(y)
    print("=========================")
    x = np.array(x)
    y = np.array(y)
    print(x)
    print(y)
    print("==========================")
    plt.figure(figsize=(6, 9))
    plt.subplot(211)
    plt.plot(x[:,0], y, 'ro')
    plt.title("TV")
    plt.grid()
    plt.subplot(212)
    plt.plot(x[:,1], y, 'g^')
    plt.title("Radio")
    plt.grid()
# =============================================================================
#     plt.subplot(313)
#     plt.plot(x[:,2], y, 'b*')
#     plt.title("Nespaper")
#     plt.grid()
# =============================================================================
    plt.tight_layout()
    plt.show()
    # feed data
    x_train, x_test, y_train, y_test = train_test_split(
            x, y, train_size=0.8, random_state=1)
    print("---------------------------")
    print(x_train)
    print(y_train)
    print("----------------------------")
    # read model
    linearRegression = LinearRegression()
    # gradien descent method to fit data parameters
    model = linearRegression.fit(x_train, y_train)
    print("=============================")
    print(model)
    print(linearRegression.coef_)
    print(linearRegression.intercept_)
    print("=============================")
    y_hat = linearRegression.predict(np.array(x_test))
    mse = np.average((y_hat - np.array(y_test)) ** 2)
    rmse = np.sqrt(mse)
    print(mse)
    print(rmse)
    
    t = np.arange(len(x_test))
    plt.plot(t, y_test, 'r-', Linewidth=2, label="real data")
    plt.plot(t, y_hat, 'g-', Linewidth=2, label="predict data")
    plt.legend(loc='upper right')
    plt.title("linear regression sales", fontsize=18)
    plt.grid()
    plt.show()

采取三个行业的特征进行分析预测:

采取前面两个线性特征比较直观的特征进行试验: