Closed mattyschell closed 1 year ago
The SQL at the center of this investigation:
Related #5
Hi @mattyschell. You may have already spotted this. In addition to the condo issue, there are also a handful of records where the block/lot are in sync, but the first digit borough code in BBL values are different. That is strange.
Thank you formerly anonymous colleague @wmc281 . Great to see you helping us out here on GitHub. I've responded to your additional issue, which I had definitely not yet spotted, with the best emoji GitHub has to offer for this.
Thanks again for these reports, feel free to open up new issues or comment on existing ones any time you like.
https://github.com/mattyschell/geodatabase-buildings/issues/14
Building bin/bbls in different boros is a known issue, resolution TBD.
I reviewed this excellent list and I see three distinct issues.
We will upgrade https://github.com/mattyschell/geodatabase-buildings-bbl-qa to flag category 3. I am closing this issue under this repository.
Thanks again @wmc281 great QA here! Stay tuned.
A colleague at NYC Emergency management reports discrepancies comparing mappluto_bbl values we produce here to actual values in MapPLUTO. Seems bad!
Quickly reviewing these I see a couple of possibilities to track down. Sometimes we have multiple condos on a single building footprint and we remove duplicates at random. Let's revisit this decision, and I use the word "decision" loosely.
We also may have cases where the condo value reported in MapPLUTO is different from the value we fetch from the live the Dept of Finance condo dataset. We chose the Dept of Finance value because it seemed like it would be the most truthy. But if the point of this repo is to provide helpful values that join to MapPLUTO then why are we populating the buildings dataset with good but ultimately not useful values?