Question

这是我的代码，我无法实现多元线性回归。我正在尝试使用梯度下降来实现它，并且在theta = theta-（learn / len（xtrain））* np.sum（xtrain *（xtrain @ theta.T-ytrain），axis = 0）中遇到错误。有人可以帮忙吗？我是python的新手。

import numpy as np
import matplotlib.pyplot as plt
import pandas as pd
from mpl_toolkits.mplot3d import Axes3D

dataset = pd.read_csv('50_Startups.csv')


x_points = dataset.iloc[:,0:3]
y_points = dataset.iloc[:,4].values
ones = np.ones([x_points.shape[0],1])
x_points = np.concatenate((ones,x_points),axis=1)


x_points = (x_points - x_points.mean())/x_points.std()
y_points = (y_points - y_points.mean())/y_points.std()


from sklearn.cross_validation import train_test_split

xtrain,xtest,ytrain,ytest = train_test_split(x_points,y_points,test_size = 1/3, random_state = 0)


theta = np.zeros(4)
learn = 0.001 # .001, .01, .1, 1
iters = 100

#computecost
def computeCost(xtrain,y,theta):
    tobesummed = np.power(((xtrain @ theta.T)-y),2)
    return np.sum(tobesummed)/(2 * len(xtrain))

#gradient descent
def gradientDescent(xtrain,ytrain,theta,iters,learn):
    cost = np.zeros(iters)
    for i in range(iters):
        z =(xtrain @ theta.T)
        theta = theta - (learn/len(xtrain)) * np.sum(xtrain * (xtrain @ theta.T - ytrain), axis=0)
        cost[i] = computeCost(xtrain, ytrain, theta)

    return theta,cost

#running the gd and cost function
g,cost = gradientDescent(xtrain,ytrain,theta,iters,learn)
print(g)

finalCost = computeCost(xtrain,ytrain,g)
print(finalCost)

操作数不能与形状（33,4）（33，）一起广播

0 个答案: