[x] basic functionality in classes PensionSim, PensionInsuranceCompany, Individual
[x] PublicOpinion as virtual Entity (reputations regenerate slowly on each step)
[x] PensionContracts
[x] Individuals search for PensionInsuranceCompany (influenced by PublicOpinion), enter PensionContracts with one of them
[x] Individuals influence PublicOpinion for PensionInsuranceCompany if something goes wrong
[x] Check that time steps are flexible
[x] Simplify Contract interface
[x] Recreate previous pensionenv implementation (without any investing)
[x] Prove learning works on this example
[ ] Add investing as a plugin
[x] Apply learner (try out something other than q-learner)
Full disclosure: actually adding/fixing a LOT of framework-related things under the hood (in fin_sim base class and fin_env interface) as this example simulation is fleshed out and tested.
Edit:
Todos:
Full disclosure: actually adding/fixing a LOT of framework-related things under the hood (in fin_sim base class and fin_env interface) as this example simulation is fleshed out and tested.