如何使用Python灵活处理不同数量变量的配置校准问题

import numpy as np

def my_func(a, b, c, d, p1, p2, p3, p4):
    """
    示例目标函数，根据实际情况替换
    """
    return a + b * p1 + c * p2 + d * p3 + p4

def create_derivative_function(mask, a, b, c, d):
    """
    动态创建导数函数
    """
    def derivative(old_params):
        derivatives = []
        param_index = 0
        all_params = [a,b,c,d] # 这里假设a,b,c,d是固定的，如果不是，需要修改
        all_params_index = 0
        for i, need_calibrate in enumerate(mask):
            if need_calibrate:
                p_value = old_params[param_index] # 从需要校准的参数列表里取值
                du = my_func(a, b, c, d, *(all_params[:i] + [p_value * 1.1] + all_params[i+1:]))
                dd = my_func(a, b, c, d, *(all_params[:i] + [p_value * 0.9] + all_params[i+1:]))
                derivative_value = (du - dd) / (0.2 * p_value)
                derivatives.append(derivative_value)
                param_index += 1 # 校准参数索引递增
            all_params_index += 1

        return np.array(derivatives)

    return derivative


def calibrating(old_params, derivative_func, learning_rate=0.01, iterations=100):
    """
    校准函数
    """
    params = np.array(old_params)
    for _ in range(iterations):
        J = derivative_func(params)
        params = params - learning_rate * J
    return params


# 示例用法
a, b, c, d = 1, 2, 3, 4  # 示例参数
p1, p2, p3, p4 = 5, 6, 7, 8  # 初始参数值

# 定义参数掩码，True表示需要校准，False表示固定
mask = [True, True, False, False]

# 创建导数函数
derivative_func = create_derivative_function(mask, a, b, c, d)

# 准备需要校准的参数的初始值
initial_guess = [p1, p2]  # 只包含需要校准的参数

# 进行校准
calibrated_params = calibrating(initial_guess, derivative_func)

print("校准后的参数:", calibrated_params)