#143: adapt sim1.m for purely forward models and remove simk.m -------------------------+-------------------------------------------------- Reporter: sebastien | Owner: sebastien Type: enhancement | Status: accepted Priority: major | Milestone: 4.3 Component: General | Version: Keywords: | -------------------------+--------------------------------------------------
Comment(by sebastien):
13:07 <michel> sebastien: I think that we need three different functions called in simul.m: sim1.m as now; one function for purelay backward models; one function for purely forward models. 13:09 <michel> sebastien: for the two purely backward/forward models, it would be useful to know the date of the last shocks. 13:13 <michel> sebastien: purely backward models: run the nonlinear solver forward, using the solution of the previous period as initial value (or a trajectory simulated with the linear approximation, if we ever introduce this option). If the residual computed with the initial value is smaller than the tolerance and that there is no shocks in the future, one can just duplicate the solution until the end of the simulation. 13:14 <michel> sebastien: purely forward models: shocks trigger precautionary behaviors, but have no effect in the future. The simulation is ran backward, but remains at the terminal condition until the date of the latest shock