Closed SPTKL closed 3 years ago
AD ideas for priority of assigning spatial attributes
1) If record gets a hit in Geosupport and and BIN or BBL is returned take all spatial attributes from Geosupport. If a spatial attribute is not returned by Geosupport do not try to backfill.
2) If a record does not get a hit in Geosupport, but we were able to create spatial data and join the dataset to a tax lot take all spatial data from PLUTO. We need to decide how to get NTA and transform censtract to proper format. A record should not join to PLUTO and not get a hit in Geosupport.
2) Take all spatial attributes via spatial join, if a record does not get a hit in Geosupport, but we were able to create spatial data.
3) Take attributes from source data if it exists.
Goals:
1) Ensure spatial attributes are consistent across the record. We'll need to run QAQC checks.
Currently, there are 41 records in FacDB where borocode <> LEFT(bbl,1)
, 67 records where borocode <> LEFT(commboard,1)
, and more; therefore, you want all spatial attributes coming from the same source.
2) Do not mix sources for spatial data.
3) Have valid spatial data attributes. Right now, there are 4 records where borocode = 0
Mapping enhancement If a point falls within the water, use a clipped boundary (i.e. clipped mappluto) to map the point to land. Points in water have caused issues with assigning the proper spatial boundary attributes.
Spatial boundary attributes are:
Address attributes are
Outlier
@AmandaDoyle we don't collect source data spatial attributes (except for boro, borocde, zipcode, which can be very wrong and inconsistent), so 4 is not possible
@SPTKL See edited comment, which was updated based on my conversation with LS. Yesterday, when I was reviewing the dataset the record GRAND FERRY PARK probably caused me the most grief and I definitely want to avoid this from happening in the new and improved FacDB
depends on:
facdb_geom
facdb_base
Possible sources for boro and borocode:
Priority?
pending ....