如何使用numpy从具有约束的损失最小化中获得r_squared？

import numpy as np

x = np.array([1, 2, 3, 4, 5])  # 输入数据
y = np.array([2, 4, 6, 8, 10])  # 目标数据

def loss_function(params):
    # 根据参数计算预测值
    y_pred = params[0] * x + params[1]
    # 计算残差平方和
    residuals = np.sum((y_pred - y) ** 2)
    # 添加约束条件
    if params[0] < 0:
        residuals += 1000  # 惩罚项，使斜率为负时损失增加
    return residuals

from scipy.optimize import minimize

# 初始参数值
initial_params = np.array([1, 1])

# 最小化损失函数
result = minimize(loss_function, initial_params)

# 获得最优参数值
optimal_params = result.x

# 根据最优参数计算预测值
y_pred = optimal_params[0] * x + optimal_params[1]

# 计算总平方和
total_sum_squares = np.sum((y - np.mean(y)) ** 2)

# 计算残差平方和
residual_sum_squares = np.sum((y - y_pred) ** 2)

# 计算R_squared
r_squared = 1 - (residual_sum_squares / total_sum_squares)