python学习笔记（28）----多元线性回归

今天又是自闭的一天，从入门到自闭，
一开始写的显示错误，然后看了半天没看出来，一开始学习率是0.001，最后调了个学习率就没事了，不知道参数怎么调，后面去看书找找

E:/pycharm_/多元线性回归.py:35: RuntimeWarning: overflow encountered in double_scalars
  k1_grad += x[j,0] * (k1 * x[j,0] + k2 * x[j,1] + k0 - y[j]) / float(len(x))
E:/pycharm_/多元线性回归.py:38: RuntimeWarning: invalid value encountered in double_scalars
  k1=k1-lr*k1_grad
nan nan nan
E:/pycharm_/多元线性回归.py:54: UserWarning: Z contains NaN values. This may result in rendering artifacts.
  ax.plot_surface(x0, x1, z)

import numpy as np
im