Page 1 of 1

Dynare v 4.0.4 crash

PostPosted: Tue Sep 01, 2009 4:57 pm
by ebaykal
Hi everybody,

I was using dynare version 3, and just downloaded version 4.0.4 for matlab. I have matlab R2008a installed on 32 bit system.

I did not have any problems in version 3 with my mod file. But when I try to run it in dynare v 4.0.4 and if the periods=100000 is declared at the beginning of the mod file, it takes forever for dynare to produce the output. If I comment out periods=100000, it works fine.

Does anybody know why it is taking forever for dynare to compute with periods declared at the beginning of the mod file?

Thanks for the help in advance.

Elif.

Re: Dynare v 4.0.4 crash

PostPosted: Wed Sep 02, 2009 8:24 am
by SébastienVillemot
Hi,

First "periods=10000;" is not a valid Dynare syntax. There is an old keyword "periods" which is now deprecated, which is used to declare the number of periods in a deterministic simulation. The right syntax does not have an equal sign, and is described there:

http://www.dynare.org/manual/re25.html

However you should rather use the "periods" option of "simul" and "stoch_simul" (see their respective meaning in the reference manual).

Note that it is perfectly normal that the computation takes more time when you increase the "periods" option: you are simply giving more work to Dynare...

Best,