Deterministic simulation solves the limit

I might ask a general question.I want to ask how big a model can Dynare solve using deterministic simulation?

I am not sure there is a good answer. The limit should depend on your machine’s memory and computing time. What did you have in mind?

I feel that when Dynare carries out deterministic simulation, the nonlinear solution often does not converge even if it is set to a large number of iterations, and most of the deterministic simulation relies on homotopy algorithm.I would like to ask again if there is a mod for the two-country model.

I doubt that it’s only about the size of the model. For example, linear equation systems can be solved easily even if they are huge. So most probably it’s about larger models featuring more nonlinearity.

I’d like to ask one more question.Do you have an example mod file for the two country model?

No, I don’t have such a mod-file

I’d like to ask one more question.I’ve seen some use of it in some papersThe setting of piecewise functions, how should this be implemented in Dynare?Should macros be used?

What do you mean with “piecewise functions”?

For example,this is a function y=f(x),when x>=0,y=b,and thenx<=0,y=a.

So you do not simply mean an occasionally binding constraint like the ZLB but really a piecewise function?

Yes.the real piecewise functions.

You could always use an external function. The problem is the solution technique you are using. Such functions are typically non-differentiable so that perturbation techniques cannot be used.