Error using extended path command

[h0ps]

Hi

I get the following error message when I'm using the extended path command. Can anyone figure it out?

Error using homotopic_steps (line 151)
extended_path::homotopy: Oups! I did my best, but I am not able to simulate this
model...

Error in extended_path (line 314)
[INFO,tmp] = homotopic_steps(endo_simul,exo_simul_1,.5,.01,pfm1);

Error in NKmodel_baseline (line 139)
extended_path([], 100);

Error in dynare (line 174)
evalin('base',fname) ;

[jpfeifer]

Please try the most recent snapshot. There was a bug in stochastic extended path with homotopy. If the problem persists, please post the mod-file

[h0ps]

Hi

I will try that. Thanks

[racket]

Hello, everyone!
I am trying to use the extended_path command (in Dynare 4.4.3), but I get the following error message:

Error using chol
Matrix must be positive definite.

Error in setup_stochastic_perfect_foresight_model_solver (line 23)
pfm.Omega = chol(pfm.Sigma,'upper'); % Sigma = Omega'*Omega

Error in extended_path (line 39)
pfm = setup_stochastic_perfect_foresight_model_solver(M_,options_,oo_,'Tensor-Gaussian-Quadrature');

Error in shock_eaa_tech (line 19126)
extended_path([], 100);

Error in dynare (line 180)
evalin('base',fname) ;

Any ideas of what might be wrong?
In this context, I just want to ask: is it possible to use the extended_path command for a deterministic model?
I am interested in somehow generating theoretical moments for the variables of the model.

Thanks a lot!

[jpfeifer]

1. Always provide files to replicate the issue
2. Why would you want to use extended path with a determinstic model? In this case, everything is known and no integrals need to be evaluated. In this case, Newton-type solvers can usually already solve your model, i.e. use simul (although it is not clear to me why someone wants moments from a deterministic model)
3. You cannot generate theoretical moments from a simulation-based solution.