[Imports] [Talk-us-newyork] [Imports-us] NYS GIS SAM Address Point Import - Status Update
Skyler Hawthorne
osm at dead10ck.com
Wed Jun 2 00:12:24 UTC 2021
Jun 1, 2021 12:54:24 Jmapb <jmapb at gmx.com>:
> Bravo, further along than I realized!
>
> My own experience with the import has been mostly in Ulster and a little
> in Greene. A couple items of possible interest around Phoenicia:
>
> https://www.openstreetmap.org/way/20216874 is "Ed West Road" (imported
> from TIGER and confirmed by survey.) The address nodes along it were
> imported with the street name "Ed Road West." This may be an error in
> the SAM data, not in your code.
>
> Along Mount Ava Maria Drive (
> https://www.openstreetmap.org/way/364548986 ), there were some buildings
> already tagged with address and some without. Those with no address were
> tagged with "addr:street=Mt Ava Maria Drive" (eg
> https://www.openstreetmap.org/way/716850634 ) but those already tagged
> with "addr:street=Mount Ava Maria Drive" seemed to be successfully
> matched and received their nysgissam:nysaddresspointid tags etc (eg
> https://www.openstreetmap.org/way/716850634 .) All of them, however, got
> the "nysgissam:review=no matching street nearby" tag. I'm guessing this
> is a quirk with the expansion of "Mt."
>
> These are easy fixes, but I thought it might be helpful to see them in
> the current state. These are the only two definitive problems I've seen
> with the SAM street names. In general, encountering "nysgissam:review=no
> matching street nearby" is a good indication of erroneous TIGER data.
Yeah, I checked the raw data for these address points, and in this case, it looks like the data itself had the abbreviation, which is unusual based on my experience with what I've seen so far. The data is definitely wrong in this case. Hopefully it's easy enough to fix these with a bulk edit in JOSM.
> Of course there are also issues around street names for numbered-route
> addresses, previously discussed and perhaps worth rehashing at some
> point.
Yeah, I've abstained from trying to solve the numbered route problem as part of this import. It's a whole separate problem, and even the existing data in OSM is very inconsistent. Even official government sources can't agree on a standard, because I don't think one exists.
I think to really fix this problem, we'll all have to agree on one way to name and tag these roads (if indeed there is any hope of consensus on this), and then embark on a statewide effort to retag all existing roads and addresses.
Down the road, if this hasn't been done yet, I'd be happy to lend my hand to a large scale automated fix of the data. I think this sounds entirely possible with Postgres.
> I'm also hoping we can arrive at some post-import QA guidelines,
> once you've finished all the counties.
Yeah, I agree, based on the questions I've gotten about the data so far, I think a guideline will help. My goal after the initial import is first to update the wiki page with some more detail about what you can expect to find in the data, how and when things conflate and when they don't, suggestions for how to fix each review tag, etc.
After that, I'll be figuring out how to do updates. The data is updated on a quarterly basis, and I'll need to figure out how to handle changes, especially to data that has been edited in the time between the initial import and when the update is performed.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.openstreetmap.org/pipermail/imports/attachments/20210602/47310fc3/attachment.htm>
More information about the Imports
mailing list