IRE_XBND problem

[Antti-L]

I have tried to specify IRE_XBND for bounding the total exports from an external region.  Let's say this external region is OPEC, and the commodity is OIL.  But likewise, one should be able to use the external region IMPEXP.

I have tried it in a Trade scenario, but I am getting a spurious "Commodity is invalid" error (Step2).  The commodity is certainly valid, because VEDA creates the TOP_IRE links e.g. TOP_IRE('IMPEXP','OIL','REG1','OIL','IMPOIL1'), and so OIL is of course a perfectly valid commodity in the region IMPEXP.  In my tests, VEDA reads the IRE_XBND parameters quite correctly, but in "Step 2" it deletes them, giving the reason "Commodity is invalid", which is not true.

Can the VEDA developers give me an example of how to specify such a bound IRE_XBND('IMPEXP', '2030','OIL','ANNUAL','EXP','UP')=12345, e.g. for the Demo7 model?

[AKanudia]

Thanks for pointing this out. It is indeed happening because Veda checks for valid region-commodity and the source list does not have external regions. We will fix this.

[Antti-L]

Thanks!  But please consider doing that for IRE_XBND only, and not for any commodity attributes in any scenarios (that would cause problems).  IRE_XBND is a rather special case.

[AKanudia]

How about ignoring cases where the region in question is an exogenous region?

[Antti-L]

> How about ignoring cases where the region in question is an exogenous region?

Ehh... sorry I am not following here, can you elaborate?  IRE_XBND is valid for external regions, but other attributes are in general not.

[Edit:] If other attributes defined for external regions would continue to be ignored (as they are now), then yes, I reckon that would work. The IRE_XBND would then however survive, because the region is specified in Other_indexes?

[AKanudia]

The check is done after resolving other_indexes, so it should be OK. I will send you an EXE to test later today.