0.引入依赖
import numpy as np
import matplotlib.pyplot as plt
1.导入数据
points = np.genfromtxt('data.txt',delimiter = ',')
//提取points中的两列数据,分别作为x,y
x = points[:,0]
y = points[:,1]
//用plt画出散点图
plt.scatter(x,y)
plt.show()
2.定义损失函数
//损失函数是系数的函数,另外还要传入数据的x,y
def compute_costs(w,b,points):
total_cost = 0
M = len(points)
//逐点计算平方损失误差,然后求平均
for i in range(M):
x = points[i,0]
y = points[i,1]
total_cost += (y-w*x-b)**2
return total_cost/M
3.定义模型超参数
alpha = 0.0001//下降的步长
init_w = 0
init_b = 0
num_iter = 10//相当于循环次数
4.定义核心梯度下降函数
def grad_desc(points, init_w, init_b, alpha, num_iter):
w = init_w
b = init_b
//定义一个list保存所有损失函数的值,用来显示下降过程
cost_list = []
for i in range(num_iter):
cost_list.append( compute_costs(w,b,points) )
w, b = step_grad_desc( w, b, alpha, points )
return [w, b, cost_list]
def step_grad_desc( current_w, current_b, alpha,points ):
sum_grad_w = 0
sum_grad_b = 0
M = len(points)
//对每个点代入公式求和
for i in range(M):
x = points[i,0]
y = points[i,1]
sum_grad_w += ( current_w * x + current_b - y ) * x
sum_grad_b += current_w * x + current_b - y
//用公式求当前梯度
grad_w = 2/M * sum_grad_w
grad_b = 2/M * sum_grad_b
//梯度下降,更新当前的w,b
updated_w = current_w - alpha * grad_w
updated_b = current_b - alpha * grad_b
return updated_w,updated_b
5.测试,运行梯度下降算法
w, b, cost_list = grad_desc(points, init_w, init_b, alpha, num_iter)
print("w is: ", w)
print("b is: ", b)
cost = compute_costs(w,b,points)
print("cost is: ", cost)
plt.plot(cost_list)
plt.show()
w is: 1.3889185526309822
b is: 0.02708960089671493
cost is: 98.70002273561147
6.画出拟合曲线
plt.scatter(x, y)
//针对每一个x,计算出预测的y的值
pred_y = w * x + b
plt.plot(x, pred_y, c = 'r')
plt.show()