Error using horzcat in deterministic simulation

Dear all,

I’m working with a model with search and matching frictions and endogenus separation (thus, I linearized the stochastic integral describing the expected productivity around the level of reservation productivity in the mod file). I was just trying a deterministic shock on the nominal interest rate to verify the dynamics of the model when I got this message:

Error using horzcat
Dimensions of matrices being concatenated are not consistent.

Error in simul (line 58)
mess = 'SIMUL: error in model specification : variable ’
M_.endo_names(find(M_.lead_lag_incidence(M_.maximum_lag+1,:)==0),:)];

Error in shocks (line 761)
simul();

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

I really do not understand where the error comes from. The strange thing I noticed is that inside the oo_ structure, exo_simul is not a 1X8 vector but a 2X8 matrix, which I can’t explain.

Does someone have any clue on why this happens? I attach the files.

Thank you!
var_param_nonlin.mod (5.93 KB)
lmr_model_nonlin.mod (7.1 KB)
steady_state.m (8.38 KB)

The crash is due to a bug in constructing the error message. It should be

[quote]PERFECT_FORESIGHT_SETUP: error in model specification : the variable(s) n_n, n_e don’t appear as current period variables.
[/quote]