The RATS Software Forum

Dear Tom,

I used successfully the @SURGibbsSetup command for a Near-VAR. Now, I would like to introduce a threshold and estimate the near-VAR for values larger and smaller than zero for a specific variable. Here the code:

===================================================================
set low = IRI<0
set high = IRI=0

...

sur(outsigma=v,smpl=low) 5
# equation1
# equation2
# equation3
# equation4
# equation5

group varmodel equation1 equation2 equation3 equation4 equation5

@SURGibbsSetup varmodel smpl=low

...

==================================================================================
I think that the way I specified the command (@SURGibbsSetup varmodel smpl=low) is not correct, since, as far as I know, @SURGibbsSetup works only with the options "start" and "end". Could you please suggest a solution to this problem?

Many thanks.

Best,
Adrian

The only place where the data enter into @SURGibbsSetup is on

cmom start end
# %rlfromtable(table)

You can add the option

option SERIES SMPL

then change the cmom to

cmom(smpl=smpl) start end
# %rlfromtable(table)

You would then execute it with something like:

@SURGibbsSetup(smpl=low,model=...)

Thanks a lot Tom, it works!

Adrian