readIEA() drop all data for USSR

[0UmfHxcvx5J7JoaOhFSs5mncnisTJJ6q]

mrcommons/R/readIEA.R

Line 25 in 8768783

data$COUNTRY <- toolCountry2isocode(data$COUNTRY, warn = FALSE) # nolint

[fbenke-pik]

@0UmfHxcvx5J7JoaOhFSs5mncnisTJJ6q Just to be sure, you want to have the USSR (aka "SUN") data removed altogether for readSource("IEA", subtype = "EnergyBalances"), right? Or do you want it to be disaggregated into SUN countries?

[0UmfHxcvx5J7JoaOhFSs5mncnisTJJ6q]

It is being dropped right now. I always assumed there was no data from the USSR in the IEA balances, because I always looked at the output of this function before. That was my mistake.
I would very much like to use disaggregated data for the USSR prior to 1990.

[fbenke-pik]

Ok, thanks for clarification. I had a closer look. There is a madrat function toolISOhistorical applied that should take care of resolving data of regions that no longer exist like USSR. Problem is, the function uses as weight the values of the "transition year", in the case of USSR 1991. Since this is NA in the sources, the historical data just gets 0s as weights and is effectively removed.

I can think of two ways to fix this and would like @LaviniaBaumstark to pick one, as she is the author of toolISOhistorical.

1. Disaggregate USSR outside the madrat helper toolISOhistorical manually
2. Enhance madrat helper toolISOhistorical to accept an optional weight

[LaviniaBaumstark]

2. sounds good, but maybe also
3. adjust toolISOhistorical to use optional other year for transition year

[fbenke-pik]

3. wont help here, because it has to be a transition year after the end of the extinct region (so you have values for all the countries the region is dissolved into). looks like data for all years after 1991 is 0 or NA for the data I want to take care of

[fbenke-pik]

besides, 3) can already be achieved by changing the transition year passed with the argument "mapping"

[fbenke-pik]

I finished implementation. @0UmfHxcvx5J7JoaOhFSs5mncnisTJJ6q

since the magclass object is really big, any idea how i can test locally before submitting the PRs for review? i.e. any variables you care about in particular for which i should check if the disaggregation worked?

[0UmfHxcvx5J7JoaOhFSs5mncnisTJJ6q]

Don't.

sbatch --wrap "Rscript -e \"devtools::load_all(<your mrcommons branch>); saveRDS(mrcommons:::readIEA(subtype = 'EnergyBalances'), 'readIEA_EnergyBalanaces.rds')" --output=log.txt --qos=priority

Then look at the results at leisure.

[fbenke-pik]

Should be solved by PR #44