收到类型错误：找不到所需的参数“对象”（位置 1）

Question

我正在尝试创建一个移动鼠标的健身房环境（显然是在 VM 中）......我对类不太了解，但是应该有一个关于 self 或其他东西的争论......？此外，任何改进将不胜感激......

这段代码基本上要在虚拟机上运行，​​所以......我试图删除这行代码，但有几行没有运行......（我在解释事情方面很糟糕）

这是代码：

class MouseEnv(Env):
    def __init__(self):
        self.ACC = 0
        self.reward = 0
        self.done = False
        self.reset()

    def step(self, action):
        try:
            self.action = action
            done = False

            if self.action == 1:
                pyautogui.click()
                self.reward += 0.2
            else:
                if self.ACC == 1:
                    self.action = min((self.action/100), 1) * 1920
                    self.prev_action = min((self.prev_action/100), 1) * 1080
                    self.reward += 0.4

                else:
                    self.ACC = 1
                    self.prev_action = self.action()
                    self.reset()
            screen = ImageGrab.grab()
            self.observation = np.array(screen)
        except:
            done = True
        return self.observation, self.reward, done, {}           
    def reset(self):
        self.observation = np.array()
        self.reward = 0
        self.done = 0
        return self.observation

和错误：

Traceback (most recent call last):
  File "<stdin>", line 1, in <module>
  File "/home/rees/.local/lib/python3.6/site-packages/gym/envs/registration.py", line 171, in make
    return registry.make(id, **kwargs)
  File "/home/rees/.local/lib/python3.6/site-packages/gym/envs/registration.py", line 123, in make
    env = spec.make(**kwargs)
  File "/home/rees/.local/lib/python3.6/site-packages/gym/envs/registration.py", line 87, in make
    env = cls(**_kwargs)
  File "/home/rees/Desktop/gym-mouse/MouseGym/envs/mouse_env.py", line 12, in __init__
    self.reset()
  File "/home/rees/Desktop/gym-mouse/MouseGym/envs/mouse_env.py", line 41, in reset
    self.observation = np.array()
TypeError: Required argument 'object' (pos 1) not found

预期结果：

我希望鼠标根据代理的输入/动作移动，并且观察是屏幕的实时视频馈送...

Answer 1

问题在于，在您的reset函数中，当您尝试self.observation使用空 numpy 数组进行初始化时，您没有将任何参数传递给np.array(). 这里有两个选项，将一个空列表传递给函数：

self.observation = np.array([])

或使用np.empty，这将创建一个空的 numpy 数组：

self.observation = np.empty(0)