You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
The point of the FramedAgentType is to make an AgentType class that has all of the model-specific logic modularized out into configurable Python functions and objects. (Eventually, these will be configured from Dolang configuration files).
Once there is properly designed FramedAgentType, it should be possible to write a generalized backwards induction solver that works on any model that is well defined using Frames.
This ticket is for that general backwards induction solver for the micro- agent case. (I.e., without market equilibria)
The text was updated successfully, but these errors were encountered:
There's now a 'block' based system for defining models. (This supersedes the FramedAgentType, which should be removed #1428)
One part of this general solver is an algorithm for solving a specific stage #1438
The micro agent general solver will require looping that stage solving equation appropriately, with a provided set of approximation and solver method parameters.
The point of the FramedAgentType is to make an AgentType class that has all of the model-specific logic modularized out into configurable Python functions and objects. (Eventually, these will be configured from Dolang configuration files).
Once there is properly designed FramedAgentType, it should be possible to write a generalized backwards induction solver that works on any model that is well defined using Frames.
This ticket is for that general backwards induction solver for the micro- agent case. (I.e., without market equilibria)
The text was updated successfully, but these errors were encountered: