Understanding special_co addresses

[planemad]

While trying to rework on the workflow for address cleanup, @Rub21 and I discovered inconsistencies in the special_co addresses that require cleanup.

A address

446 Grandview Ave was marked for removal. Another address 2049 Himrod St was located for the same building.

http://maps.nyc.gov/

• Both 446 Grandview Ave and 2049 Himrod St point to the exact same building

Google Streetview

• In streetview we found that 20-49 is the address of the store whose entrance is on Himrod St.
• The entrance to the building itself is on the side and is 446 on Grandview Ave

  B address

This is a cluster of duplicate addresses but only 3 are marked as B code addresses (red) in Queens.

• The street address Prospect Ave does not match the actual street name 60th Place
• This address is valid on maps.nyc.in
• The other duplicates follow the same pattern but do not appear to be part of the code B address set

  Conclusion

It is unclear on what the exact status of these special_co A and B addresses are. They all seem to be valid addresses on http://maps.nyc.gov, and in some cases even exist physically as in the first case, so it will be incorrect to remove them.

@lxbarth is there any further documentation or discussion of these addresses and are we 100% sure they need to be cleaned up?

[lxbarth]

@planemad - take a look at back story here https://github.com/osmlab/nycbuildings/issues/35

[planemad]

Thanks @lxbarth

Capturing from Pg12 of http://www.nyc.gov/html/dcp/pdf/bytes/lion_metadata.pdf:

VALUE A Alternate Address Range: Alternative address ranges for the same street name. This can occur where buildings have been renumbered; old numbers will sometimes remain in use. For example, such usage is common in some Queens neighborhoods, including Far Rockaway, Douglaston, Forest Hills and Ridgewood, where non-hyphenated addresses have been replaced by hyphenated addresses.

VALUE B Alternative Street Names: Alternative street names that cannot be handled in the usual way.

[lxbarth]

We're all set here - tbc on #139