打印维度是（100，2），函数内运行是（100，1）？哪里错了？

Question

from playML.LinearRegression import LinearRegression
import numpy as np
import matplotlib.pyplot as plt

if __name__ == "__main__":

    x = np.random.random(size=100)
    y = x * 3. + 4. + np.random.normal(size=100)
    # plt.scatter(x,y)
    # plt.show()
    X = x.reshape(-1,1)
    X_b = np.hstack([np.ones((len(y),1)),X])
    print(X_b.shape)
    lin_reg = LinearRegression()
    lin_reg.fit_gd(X,y)

def fit_gd(self, X, y, eta=0.01, n_iters=1e4):
        assert X.shape[0] == y.shape[0], \
             "The size of X_train and y_train must be same!"
        X_b = np.hstack([np.ones((len(y), 1)),X])
        initial_theta = np.zeros(X_b.shape[1])

        def J(theta, X_b, y):
            try:
                return (y -X_b.dot(theta)) ** 2 / len(y)
            except:
                return float('inf')

        def dJ(theta, X_b,y):
            # res = np.empty(len(theta))
            # res[0] = np.sum(X_b.dot(theta) - y)
            # for i in range(1,len(theta)):
            #     res[i] = (X_b.dot(theta - y)).dot(X_b[:,i])
            return X_b.T.dot(X_b.dot(theta) - y) * 2 / len(theta)
        
        def gradient_descend(X_b,y,initial_theta,eta=0.01,n_iters=1e4,epsilon=1e-8):
            theta = initial_theta
            i_iters = 0
            while i_iters < n_iters:
                gradient = dJ(theta,X_b,y)
                last_theta = theta
                theta = theta - eta * gradient
                if abs(J(last_theta,X_b,y) - J(theta,X_b,y)) < epsilon:
                    break
                i_iters += 1
            return theta

        self._theta = gradient_descend(X,y,initial_theta,eta,n_iters=1e4,epsilon=1e-8)
        self.interception_ = self._theta[0]
        self.coef_ = self._theta[1:]
        return self

运行结果：
/Users/jason/anaconda3/bin/python3.7 /Users/jason/Desktop/MarchineLearning/main-gradient-descendent.py
Traceback (most recent call last):
File “/Users/jason/Desktop/MarchineLearning/main-gradient-descendent.py”, line 15, in
lin_reg.fit_gd(X,y)
File “/Users/jason/Desktop/MarchineLearning/playML/LinearRegression.py”, line 52, in fit_gd
self._theta = gradient_descend(X,y,initial_theta,eta,n_iters=1e4,epsilon=1e-8)
File “/Users/jason/Desktop/MarchineLearning/playML/LinearRegression.py”, line 44, in gradient_descend
gradient = dJ(theta,X_b,y)
File “/Users/jason/Desktop/MarchineLearning/playML/LinearRegression.py”, line 38, in dJ
return X_b.T.dot(X_b.dot(theta) - y) * 2 / len(theta)
ValueError: shapes (100,1) and (2,) not aligned: 1 (dim 1) != 2 (dim 0)
**(100, 2)**这是打印的维度

Answer 1

liuyubobobo 2020-02-09 02:31:08

你传的是 X 不是 X_b？

• 提问者 Jason_chen0755 #1

  X_b 封装在fit_gd里面; dJ函数里应该传X而不是X_b？

• 提问者 Jason_chen0755 #2

  老师，按我的代码，为什么我在 dJ 函数里传X，而不是X_b，反而得到正确的打印结果？

• liuyubobobo 回复 提问者 Jason_chen0755 #3

  你需要下载课程的官方代码，在你的环境下运行一下，看是不是也有问题？如果没有问题，请仔细调试，和自己的程序比对一下，看一下问题在哪里。你这样贴代码，我无法给你调试，请谅解。加油！：）

点开查看后面2条评论