Python实现最简单的线性回归算法

#coding:utf-8
from math import sqrt
import matplotlib.pyplot as plt

dataset = [[1.2,1.1],[2.4,3.5],[4.1,3.2],[3.4,2.8],[5,5.4]]
x = [row[0] for row in dataset]
y = [row[1] for row in dataset]
pre_y = list()#画图用

#计算均值
def mean(values):
    return sum(values)/float(len(values))

#计算x和y的协方差函数
def covariance(x,mean_x,y,mean_y):
    cover = 0
    for i in range(len(x)):
        cover += (x[i] - mean_x)*(y[i] - mean_y)
    return cover

#计算方差
def variance(values,mean):
    return sum([(x-mean)**2 for x in values])

#计算回归系数
def coefficients(dataset):
    x_mean,y_mean = mean(x),mean(y)
    w1 = covariance(x,x_mean,y,y_mean) / variance(x,x_mean) #公式推导出来的
    w0 = y_mean - w1 * x_mean #公式推导出来的
    return (w0,w1)

#利用“均方差误差公式RMSE”求误差
def rmse_metric(actual,predicted):
    sum_error  = 0.0
    for i in range(len(actual)):
        predicted_error = predicted[i] - actual[i]
        sum_error += (predicted_error**2)
    mean_error = sum_error / float(len(actual))
    return sqrt(mean_error)

#构建简单的线性回归
def simple_linear_regression(train,test):
    predictions = list()
    w0,w1 = coefficients(train)
    for row in test:
        y_model = w1 *row[0] + w0
        predictions.append(y_model)
    return predictions

#评估算法数据准备以及协调
def evaluate_algorithm(dataset, alogrithm):
    test_set = list()
    for row in dataset:
        row_copy = list(row)
        row_copy[-1] = None
        test_set.append(row_copy)
    predicted = alogrithm(dataset,test_set)
    for val in predicted:
        pre_y.append(val)
        print('%.3f\t'%(val))
    actual = [row[-1] for row in dataset]
    rmse = rmse_metric(actual,predicted)
    return rmse

#返回RMSE
rmse = evaluate_algorithm(dataset,simple_linear_regression)
print('RMSE:%.3f'%(rmse))

#画出散点图
x = [row[0] for row in dataset]
y = [row[1] for row in dataset]
print(len(pre_y))
plt.axis([0,6,0,6])
plt.plot(x,y,'bs')
plt.plot(x,pre_y,'ro')
plt.plot(x,pre_y,'r-')
plt.grid()
plt.show()

【结果】