Closed m-jamieson closed 4 years ago
Hold off on merging this - I need to address one more issue with units. Right now, I think the dictionary processes don't properly account for the changing units.
Alright - this is ready to go now. Along with the issues 66 and 81, this fixes a couple other things that I found while resolving these:
@jump2conclusionsmatt we are trying to test the new flow mappings, but are having some issues. Did you commit coal_mining_lci.csv
I had not committed the new coal mining data yet. Should be up now. Checking to see if I missed any other data files.
With commit #505aae9 I modified setup.py to force install of a specific release of fedelemflowlist upon pip install. We should make sure that's the release that contains all the necessary mappings. @jump2conclusionsmatt @bl-young
We are currently testing the mapping file from this branch so we should have a final version to push to FEDEFL in the next couple of days (cc @hottleta)
I'm good with merging in this major update to the upstream data when you've checked it
This pull request is to resolve issues 66 and 81 - at least on the eLCI side. The function that maps emissions from the various NETL inventories to the federal elementary flow list (fedefl) will no use the mappings from the fedefl rather than the csv file previously contained in the data directory. This should close issue 66.
There is also a fix to issue 81 - flows with mapped UUIDs that do no exist are dropped from the final mapped dataframe to avoid populating openLCA with exchanges that will result in errors. The ultimate resolution to the problem will require updated mappings in the fedefl, so that no flows are dropped.