Thanks Johannes. I have fixed the bug in master. It appeared on July 11th only and concerns only the computation of the rank when using check. The fix will appear in Sunday morning's snapshot. All the best, Michel On 07/13/2012 02:11 PM, Johannes Pfeifer wrote:

Here are the two files. As its work in progress they still look very messy.

*Von:*dev-bounces@dynare.org [mailto:dev-bounces@dynare.org] *Im Auftrag von *Michel Juillard *Gesendet:* Freitag, 13. Juli 2012 13:52 *An:* List for Dynare developers *Betreff:* Re: [DynareDev] Rank condition differnce in 4.2.5 versus current snapshot

That looks like a bug in 4.3 and snapshot Can you send the example, please?

Michel Juillard

On Jul 13, 2012, at 13:03, "Johannes Pfeifer " <jpfeifer@gmx.de <mailto:jpfeifer@gmx.de>> wrote:

Dear Michel,

I am working on a model where both Dynare 4.2.5 and the most recent snapshot tell me the BK conditions are verified. However, 4.2.5 says the rank condition is satisfied, while the most recent snapshot says it is not satisfied.

The reason it that the matrix

w(1:nyf,nd-nyf+1:end)

in dyn_first_order_solver.m contains two rows of 0 that were not there in dr1.m in 4.2.5. Is there any explanation for this difference and which version should I trust? The IRFs are the same in both versions and model_diagnostics does not report any problems.

Best,

Johannes

_______________________________________________ Dev mailing list Dev@dynare.org <mailto:Dev@dynare.org> https://www.dynare.org/cgi-bin/mailman/listinfo/dev