Dynare 4.3.0 vs 4.1.1


I post this message on the off-chance that someone has an immediate answer without needing to go through the files:

Would there be any obvious reason why a mod file which ran perfectly through complete Bayesian estimation under 4.1.1 would fail under 4.3.0 (in this case with the message that “The steady state contains NaN or Inf”)?

Many thanks

If you post the mod, the steadystate and the datafile (or send them) we might be able to find out.

Thanks for the offer Johannes, but I don’t think it’s worth your effort - I have the earlier results which are perfectly useable; it was just a question of curiosity.