路演 (1)

import math
import numpy as np
from numpy import genfromtxt
import matplotlib.pyplot as plt
from mpl_toolkits.mplot3d import Axes3D

# 导入数据
data = genfromtxt("Delivery.csv", delimiter=",")
x_data = data[1:, :2]  # 选择前两列作为特征
y_data = data[1:, 2]    # 选择第三列作为目标变量

# 检查特征数量
num_features = x_data.shape[1]

# 初始化参数
weights = np.zeros(num_features + 1)  # +1 for w0
alpha = 0.005
epsilon = 1e-4
error = float('inf')  # 初始化为无穷大

m = len(x_data)

while error > epsilon:
    prediction = np.dot(x_data, weights[1:]) + weights[0]  # 计算预测

    error_i = prediction - y_data
    grad_w = np.dot(x_data.T, error_i) / m  # 计算梯度
    weights[1:] -= alpha * grad_w
    weights[0] -= alpha * np.mean(error_i)  # 更新 w0

    # 计算误差
    error = np.linalg.norm(grad_w)  # 计算误差
    print(weights, error)

# 绘制三维图
ax = plt.figure().add_subplot(111, projection='3d')
ax.scatter(x_data[:, 0], x_data[:, 1], y_data, c='r', marker='o', s=100)

x0 = np.unique(x_data[:, 0])
x1 = np.unique(x_data[:, 1])
X0, X1 = np.meshgrid(x0, x1)  # 使用唯一值生成网格
Z = weights[0] + weights[1] * X0 + weights[2] * X1

ax.plot_surface(X0, X1, Z, alpha=0.5)
ax.set_xlabel('Miles')  # 根据需要修改标签
ax.set_ylabel('Number of Deliveries')  # 根据需要修改标签
ax.set_zlabel('Time')  # 根据需要修改标签
plt.show()