Eia923 2022 final release q4 update Nov 21

[robertozanchi]

Update in progress for eia923 Aug 2023 data

[aesharpe]

There is a weird thing happening here where the rows for the plants_eia860 table go down by a small fraction. I think this has happened before, but it's such a small amount of rows that we normally ignore it. But it's so strange/annoying. Just flagging. Might poke at it again for a few hours. I wonder if it's related to #2978

For reference, the missing plants are:
plant_id_eia: 60325, report_date: 2023-01-01
plant_id_eia: 60018, report_date: 2023-01-01
plant_id_eia: 60018, report_date: 2022-01-01

[codecov]

Codecov Report

All modified and coverable lines are covered by tests ✅

Comparison is base (9b8cfb2) 88.7% compared to head (2994731) 88.7%.
Report is 3 commits behind head on dev.

Additional details and impacted files

@@          Coverage Diff          @@
##             dev   #3073   +/-   ##
=====================================
  Coverage   88.7%   88.7%           
=====================================
  Files         89      89           
  Lines      11018   11018           
=====================================
  Hits        9778    9778           
  Misses      1240    1240           

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

[cmgosnell]

wow this looked surprisingly straightforward!! thanks @robertozanchi and @aesharpe !

I have one non-blocking question about a deletion. maybe it should be blocking bc it doesn't seem like it should be deleted.

but everything else looks great! I'm glad to see that the column maps didn't need to be changed for either the early->final update or the new monthly update!

[cmgosnell]

what is this deletion about??

[aesharpe]

This deletion was an effort to fix test failures resulting from the fact that this PR is from a forked repo. But those tests still fail due to permissions anyways, so idk whether it's worth keeping this out or not. @bendnorman @jdangerx do you have thoughts?

[bendnorman]

Ah good point. Forgot that the GCS cache sync needs credentials 🤦 . I think we should still remove this because the default behavior uses the GITREF that started the action.

[bendnorman]

@jdangerx any idea why pytest.yml uses fetch-depth: 2 instead of the default fetch-depth: 1?

[jdangerx]

I bet there's something somewhere in our tox-pytest that cares about "what has changed since the previous commit." Not sure what that would be, though.

[jdangerx]

I think it's worth keeping it out, but point taken that the GCS credentials are obviously not going to be available in fork PRs. If we get more time to think about contributor workflows in the future we should think a bit harder about how we want our CI checks to go.

[zaneselvans]

In the update-conda-lockfile action there's a check for one of the actions that it's not on a forked repo. We could add that where appropriate to skip things that should run from outside the org.

[aesharpe]

Added a lil more with #3100