梯度速降法

def gradient_descent(derivative_func, initial_guess, multiplier=0.02, precision=0.001, max_iter=300):
    new_x = initial_guess
    x_list = [new_x]
    slope_list = [derivative_func(new_x)]

    for n in range(max_iter):
        previous_x = new_x
        gradient = derivative_func(previous_x)
        new_x = previous_x - multiplier * gradient

        step_size = abs(new_x - previous_x)
        x_list.append(new_x)
        slope_list.append(gradient)

        if step_size < precision:
            break

    return new_x, x_list, slope_list