# Feature Engineering To add multiple transformations to the observations of your environment, follow the following steps: 1. Subclass `hearts_gym.envs.ObsTransform`. 2. Implement a `transform` method with the same signature as given by the superclass. 3. Add the newly implemented class to the list of transformations `obs_transforms` in `configuration.py`, for example like this: ```python # Note that we are adding instances of the class. obs_transforms: List[ObsTransform] = [ MyObsTransform(), ] ``` 4. Adjust the first definition of the `obs_space` variable in the `HeartsEnv.__init__` method according to the observation space you receive after running through all transformations. [Look through the `gym.spaces` module](https://github.com/openai/gym/tree/master/gym/spaces) for help with this step. The transformations in `obs_transforms` will be applied in order of appearing in the list.