The RATS Software Forum

I'm using RATS V 8. I've just found that when I attempt to use the Stamp diagnostics procedure, I get the following message:


## SX11. Identifier NCORRS is Not Recognizable. Incorrect Option Field or Parameter Order?
>>>>number=ncorrs,<<<<
## CP24. PROCEDURE/FUNCTION STAMPDIAGS compiled with errors. Please correct

This error message occurs for programmes I've written which worked in the past with V 7.3 and with your programme ss_3_diagnostics.prg.

I thought I might have a corrupted procedure so I downloaded the current version on your website. Use of this version produced the same error message. I have tried specifying ncorrs as an option (as in the past) and without ncorrs (as in your programme). The problem appears to be with the Stampdiags procedure as Regcorrs will run correctly when the number option is specified independently of the Stamp diagnostics procedure. There are no problems with my programmes when I comment-out the Stamp diagnostics procedure.

That sounds like it's pulling in the wrong stampdiag file. If you have an older stampdiag in the same directory as the ss_... program, it will find that one before the one with the other RATS procedures.

No that isn't the problem as the procedure is in one place only and definitely not in the same folder as the programme (supplied by you) referred to in my earlier communication. As I said, I downloaded the latest version just in case it had been corrupted. For some reason ncorrs isn't being picked up. The odd thing is that the error message is being generated in programmes that have been running with this procedure ever since you supplied it.

I've just rerun the programme ss_3_3_diagnostics and got this message:

## CP24. PROCEDURE/FUNCTION STAMPDIAGS compiled with errors. Please correct
## SX11. Identifier NH is Not Recognizable. Incorrect Option Field or Parameter Order?
>>>>noprint,n=nh)<<<<

And then I reran the programme immediately. This time there was no apparent error, with following output:


State Space Model Diagnostics
Statistic Sig. Level
Q(19-1) 15.52 0.6262
Normality 0.12 0.9441
H(33) 0.61 0.1650

Just ran the programme again and got the original error message, namely, that identifier ncorrs was not recognisable.

Can you email us your program and data files (at support@estima.com) so we can try it out?

Thanks,
Tom Maycock
Estima

You have one of the programmes and data already - it's one of yours from the state space online course (see my previous messages). I'll email them anyway as they might be corrupt or, possibly, it's a problem with the V8 build that you sent me when I raised the issue of the failure of the original build to pick up models to determine an historical decomposition.