简单线性回归的实现、向量化

import numpy as np
import matplotlib.pyplot as plt

x=np.array([1.,2.,3.,4.,5.])
y=np.array([1.,3.,2.,4.,5.])

plt.scatter(x,y)
plt.axis([0,6,0,6])
plt.show()

$a=\frac{{\sum }_{i=1}^m(x^{(i)}-\bar{x})(y^{(i)}-\bar{y})}{{\sum }_{i=1}^m(x^{(i)}-\bar{x}{)}^2}$        $b=\bar{y}-a\bar{x}$

x_mean = np.mean(x)
y_mean = np.mean(y)
num = 0.0
d = 0.0
for x_i,y_i in zip(x,y):
    num += (x_i - x_mean) * (y_i - y_mean)
    d += (x_i - x_mean) ** 2

a = num / d
b = y_mean - a * x_mean

a
运行结果：0.8

b
运行结果：0.39999999999999947

y_hat = a * x + b

plt.scatter(x,y)//散点图
plt.plot(x,y_hat,color='r')//直线
plt.axis([0,6,0,6])
plt.show()

x_predict = 6
y_predict = a * x_predict + b

y_predict

输出结果：5.2

 

向量化
仅在在上述简单线性回归的步骤中进行修改，将

num = 0.0
d = 0.0
for x_i,y_i in zip(x,y):
    num += (x_i - x_mean) * (y_i - y_mean)
    d += (x_i - x_mean) ** 2

替换成

num = (x - x_mean).dot(y - y_mean)//把num和d都看成是两个向量的点乘
d = (x - x_mean).dot(x - x_mean)

得到的结果与之前一致，且向量化运算的方法使性能得到了提升