Rollout for RLDomain

[g-poveda]

I implemented a pure RLDomain
and just want to run a rollout from it (given an initial state given as input)
however the "domain.set_memory(from_memory)" fails because there is no set_memory for Environment domains such as RL ones, what would be the right way to circumvent this ?

[neo-alex]

Hello,
The ususal assumption in RL is that you are not necessarily able to initialize the environment from a given state (like in real life) - only an argument-free reset() is generally provided. In your case, you could maybe create a custom Domain inheriting from Simulation (which has a set_memory(...) method) rather than Environment (the one used generally for RL) for the dynamics characteristic.
Does that work out for you?

[g-poveda]

Yes thank you