二元线性回归的代码实现（python）

数据如下：

代码如下：

import copy
import matplotlib.pyplot as plt
import numpy as np
import pandas as pd

# 读取excel 并将其向量化为训练数据集x y
data = pd.read_excel("D:\python\python_for_beginner\成本产量数据表.xlsx", head=None, index_col=None, usecols='B' )
x_train = np.asarray(data.stack())
print(x_train)
data = pd.read_excel("D:\python\python_for_beginner\成本产量数据表.xlsx", head=None, index_col=None, usecols='C' )
y_train = np.asarray(data.stack())
print(y_train)

# 开始实现线性回归
# 函数准备


# 线性拟合函数 返回函数值列表
def computer_model_output(x, w, b):
    m = x.shape[0]
    f_wb = np.zeros(m)
    for i in range(m):
        f_wb[i] = w * x[i] + b

    return f_wb


# 成本函数 返回总偏差
def compute_cost(x, y, w, b):
    m = x.shape[0]
    total_cost =0
    for i in range(m):
        total_cost = total_cost + ((w * x[i] + b )-y[i]) ** 2

    total_cost = (1/(2*m)) * total_cost
    return total_cost


# 梯度计算函数 返回成本函数分别对w，b的偏导
def compute_gradient(x, y, w, b):
    m = x.shape[0]
    dj_dw = 0
    dj_db = 0

    for i in range(m):
        f_wb = w * x[i] +b
        dj_dw_i = (f_wb - y[i]) * x[i]
        dj_db_i = f_wb - y[i]
        dj_dw += dj_dw_i
        dj_db += dj_db_i

    dj_dw = (1/m) * dj_dw
    dj_db = (1/m) * dj_db
    return dj_dw, dj_db


# 梯度下降函数 返回__次回归后w,b的值和此处梯度
def gradient_descent(x, y, w, b, alpha1, alpha2, iter_num, compute_gradient):
    w = copy.deepcopy(w)
    while iter_num:
        dj_dw, dj_db = compute_gradient(x, y, w, b)
        w = w - dj_dw * alpha1
        b = b - dj_db * alpha2
        iter_num -= 1

    return w, b, compute_gradient(x, y, w, b)


# 学习率设置
tmp_alpha1 = 0.00000005
tmp_alpha2 = 0.00009
# 迭代次数
iter_num = 500000
w_final, b_final, dj_dw = gradient_descent(x_train, y_train, 0, 0, tmp_alpha1, tmp_alpha2, iter_num, compute_gradient)
print(dj_dw)

# 拟合出的线性函数y值
tmp_f_wb = computer_model_output(x_train, w_final, b_final)
print(w_final)

# 使中文正常显示
plt.rcParams['font.sans-serif'] = ['SimHei']
plt.rcParams['axes.unicode_minus'] = False


# 绘制出拟合曲线 和 训练集数据点
plt.plot(x_train, tmp_f_wb, c='b', label='Our Prediction')
plt.scatter(x_train, y_train, c='r', marker='x', s=50, label='Actual Value')
plt.title("成本产量关系预测图")
plt.xlabel("人工成本费（元）")
plt.ylabel("产量（公斤）")

# 显示
plt.legend()
plt.show()

运行结果如下：

为检测是否正常导入数据，梯度下降最终效果，下列输出分别为训练集数据x、训练集数据y、最终梯度（dj_dw, dj_db)、最终w。

拟合曲线如下（采用不同的学习率和迭代次数最终曲线会有不同，此代码为更好地拟合曲线采用了alpha1、alpha2两个学习率）：

希望内容对你有所帮助！也期待大佬们的纠正！