利用梯度下降实现线性拟合，效果和sklearn LinearRegression()差不多。
学习率初始设置为0.1结果算高的，直接让我的参数变成了nan。（体会到了飞出去的感觉）
把学习率调小了之后就正常了

# coding=utf-8
import matplotlib.pyplot as plt
import numpy as np
if __name__ == '__main__':
    X =[12.46, 0.25, 5.22, 11.3, 6.81, 4.59, 0.66, 14.53, 15.49, 14.43,
        2.19, 1.35, 10.02, 12.93, 5.93, 2.92, 12.81, 4.88, 13.11, 5.8]
    Y =[29.01, 4.7, 22.33, 24.99, 18.85, 14.89, 10.58, 36.84, 42.36, 39.73,
        11.92, 7.45, 22.9, 36.62, 16.04, 16.56, 31.55, 20.04, 35.26, 23.59]
    w=1
    b=0.1
    a=0.01 # 学习率
    w_temp=-100
    b_temp=-100
    wchange = 100
    bchange = 100
    while abs(wchange)>1e-12 and abs(bchange)>1e-12:
        print(wchange)
        wchange=0
        bchange=0
        for i in range(len(X)):
            wchange+=(w*X[i]+b-Y[i])*X[i]
            bchange+=w*X[i]+b-Y[i]
        wchange/=len(X)
        bchange /= len(X)
        w_temp=w-a*wchange
        b_temp=b-a*bchange
        w=w_temp
        b=b_temp
    print("y=%.4f*x+%.4f" % (w, b))



    # 简单画图显示
    plt.scatter(X, Y, c="blue")
    Y_predict=[]
    for i in range(len(X)):
        Y_predict.append(w*X[i]+b)
    plt.plot(X, Y_predict, c="red")
    plt.show()
    print(Y_predict)