Class RewardWrapper<TObs, TAct>
Transforms every reward Step(TAct) returns through Reward(double).
public abstract class RewardWrapper<TObs, TAct> : Wrapper<TObs, TAct, TObs, TAct>, IDynamicAttributes, IWrapper, IEnv, IDisposable
Type Parameters
TObsThe observation type, passed through.
TActThe action type, passed through.
- Implements
- Derived
- Inherited Members
- Extension Methods
Remarks
Reference: gymnasium/core.py.
Constructors
RewardWrapper(Env<TObs, TAct>)
Wraps env.
protected RewardWrapper(Env<TObs, TAct> env)
Parameters
envEnv<TObs, TAct>The env whose rewards are transformed.
Exceptions
Methods
Reward(double)
The transformation (reward(reward)).
public abstract double Reward(double reward)
Parameters
rewarddoubleThe env's reward.
Returns
- double
The transformed reward.
Step(TAct)
Steps the env and returns its transition with the reward transformed.
public override StepResult<TObs> Step(TAct action)
Parameters
actionTActThe action, passed on.
Returns
- StepResult<TObs>
(observation, self.reward(reward), terminated, truncated, info).