Closed AmandaDoyle closed 1 year ago
Otherwise this makes sense to me
@AmandaDoyle I'm seeing more uses of the to-be-deprecated pluto_input_research
table (example in corr_communitydistricts.sql
). but this is a WIP so no worries if those are todos!
and maybe these changes be tested later by building on this branch and comparing to 22v3.1?
Latest changes do the following
pluto_corrections
pluto_changes_applied
and pluto_changes_not_applied
just kicked off a build to test the changes
if it works and the output is identical to 22v3.1 from main
branch, would you wanna delete the commented out code? should be easy enough to revert changes via git/github if we have to
@AmandaDoyle
there's a QAQC app error since an expected pluto_corrections
export doesn't exist. I opened an issue for it and can work on it asap
we probably already mentioned this but this error is reminding me that the new research/corrections/change process maybe won't be implemented until 23v1. does that sounds right?
@damonmcc thanks for flagging. Working through issue #251 would be good, but if there implications to previous versions of PLUTO (i.e. QAQC reports of 22v2 wouldn't work), perhaps we don't rename the file, or scope out what needs to be changed?
That's right, that the earliest this change would be implemented is for 23v1, so this isn't pressing.
It seems like the the issue to get the right corrections would need to be fixed in this function function. could we just implement an if/else statement in this function? might be good to talk this through
I'll compare 22v3.1
pluto_changes_applied.csv
generated by this branch to 22v3.1
pluto_corrections_applied.csv
generated on main
can use 22v3
pluto_corrections_applied.csv
as a baseline file
@AmandaDoyle @mbh329
When comparing 22v3 to 22v3.1 applied changes, I'm seeing only one type of difference in applied changes:
field, reason = lotarea, Zero lot area
changes, the version has changed from 22v3 to 22v3.1When comparing 22v3.1 main branch to 22v3.1 this branch applied changes, I'm seeing many types of differences in applied changes:
Zero lot area
changes now have a version of 22v3.1 (maybe because it's a code-based change?)Normalized owner name
changes from pluto_input_research.csv
that were never applied are now being applied (e.g. ,ownername,NYC DOT,NYC DEPARTMENT OF TRANSPORTATION,1,Normalized owner name,19v2
a concern: I'm seeing an older version change for the same BBL/field beat a newer one when they're both on pluto_input_research. it's the changes for 1019570140,cd
. they're actually the same change but differ in reason and version
check to do: compare non-programatic field changes from applied.csv
+ not_applied.csv
and ensure they match those in pluto_unput_research.csv
lotarea
, ownername
so far seeing
input_research
than in combined change filesownername
values. see https://github.com/NYCPlanning/db-pluto/issues/429. unlikely to be caused by the code changes in this PR@damonmcc and @mbh329
pluto_input_research.csv
so that there are not duplicate corrections. BBL and field should be unique in that table.input_research
may have more table is because applied.csv
+ not_applied.csv
are limited to BBLs that currently exist in PLUTO. Perhaps we change not_applied.csv
to include bbls that are no longer in PLUTO, so that this file can be used to cleanup not_applied.csv
, or?I don't see anything that need to be changed on the code side before merging this in but before merging this in I'd like to check with LS because this will have implications on how we communicate how we make changes and the scope of changes, and that she's prepared for that for the upcoming 23v1 build.
@damonmcc and @mbh329 LS is okay merging this in and having these changes be applied to 23v1. I need you to verify that the changes on code and outputs are good to go and pass your review. Reminder that this is a maintenance item this sprint.
Addresses issue #424