I am working on a project where I have a Gym environment with a discrete action space and a continuous observation space similar to the CartPole environment from Gym. I am trying to convert the discrete action space to continuous so that the environment will work with model-based RL algos that only work with continuous actions and observations. What is the way to do this? Thank you in advance.

class SimpleEnv(gym.Env):
    def __init__(self):
        super(SimpleEnv, self).__init__()
        self.action_space = spaces.Discrete(3)
        self.observation_space = spaces.Box(low=-1, high=1, dtype=np.float32)
        self.state = np.zeros(4, dtype=np.float32)

    def reset(self):
        self.state = np.random.uniform(low=-0.5, high=0.5, size=(1,))
        return self.state

    def step(self, action):
        # Simple transition logic: modify state based on action
        if action == 0:
            self.state += 0.1
        elif action == 1:
            self.state -= 0.1
        elif action == 2:
            self.state = -self.state
        
        reward = -np.linalg.norm(self.state)

        done = np.linalg.norm(self.state) < 0.1

        info = {}
        return self.state, reward, done, info