Hello,
I am trying to migrate a model to VEDA2.0, and am struggling with getting the time periods to agree with the VEDA1.4 model. In particular, I do not understand what I need to set the ~defaultyear to be.
My data is as follows.
~StartYear
2015
~ActivePDef
2015-2050
~TimePeriods
2015-2050
1
1
1
5
5
5
5
5
5
5
What should I set ~defaultyear to be, in order to get the years to be
2015, 2016, 2017, 2020, 2025, 2030, 2035, 2040, 2045, 2050?
The old version used to take the first milestoneyear as the default year, which is not good because you could end up with a different default years with different period specifications. Veda2.0 takes the StartYear as default, if ~defaultyear is not provided.
Thank you Amit. I have set ~defaultyear = 2015. When I solve and look at GDXDiff, the AllYear, and B variables appear to differ in the model runs ("Keys are Different")
Is there any other parameter that could contribute to this difference?
PS. Note, I have many other differences in the GDXDiff, but I was thinking that getting the years to match should be one of the first steps
PPS. I should say, that I am new to TIMES/VEDA, so I am still trying to understand all the instructions on the VEDA migration
I suggest you look at the other parameters. There were some superfluous declarations for years in the old version that have been cleaned up.
But I must say that the migration process can be quite overwhelming for someone who is new to TIMES/VEDA. Please see if you can work with someone who is very familiar with the model under old Veda.
16-04-2021, 04:42 PM (This post was last modified: 16-04-2021, 04:46 PM by Antti-L.)
If the B(t) parameters are different, although the TimePeriods definition is the same, I think that is surprising, and I cannot see how that could result from other parameters. The first one should be B(t1) = StartYear, and all the rest are directly derived from TimePeriods. If there is a difference, it would affect model results.
A difference in the set Allyears would mean that the TIMES BOTIME and/or EOTIME controls are different. I think VEDA sets BOTIME according to the earliest DATAYEAR, and so it could be affected by some past years parameters.
(16-04-2021, 04:42 PM)Antti-L Wrote: If the B(t) parameters are different, although the TimePeriods definition is the same, I think that is surprising, and I cannot see how that could result from other parameters. The first one should be B(t1) = StartYear, and all the rest are directly derived from TimePeriods. If there is a difference, it would affect model results.
A difference in the set Allyears would mean that the TIMES BOTIME and/or EOTIME controls are different. I think VEDA sets BOTIME according to the earliest DATAYEAR, and so it could be affected by some past years parameters.
Thanks. The TimePeriods definition in SysSettings is identical. I set ~defaultyear = 2015 in VEDA2.
B(t1) = 2014 in both cases. However, the length(B_VEDA1.4) = 11 and length(B_VEDA2.0) = 12. Any thoughts as to where this could arise from?
Thanks for explaining the difference in AllYears. These are different lengths (VEDA1.4 starting in 1970, VEDA2.0 starting in 1980). I will have a look to try to find the earliest DATAYEAR that could account for this.
I attach a document with screenshots of the differences in the year data
Ok, I can see that the B(..) parameter under VEDA2.0 for some reason appears to be quite incorrect with respect to the TimePeriods. @Amit should be able to explain why there is such an odd difference...
(16-04-2021, 07:03 PM)Antti-L Wrote: Ok, I can see that the B(..) parameter under VEDA2.0 for some reason appears to be quite incorrect with respect to the TimePeriods. @Amit should be able to explain why there is such an odd difference...
I found the issue for the discrepancy in B(). I hadn't noticed the option to select specific TimePeriods definition when creating the saved case, and it had defaulted to one of the other definitions, rather than the ~ActivePDef
Selecting the same TimePeriod as ~ActivePDef sorts the B() issue.
Thank you for your help. I will now look at the other issues
