-
-
Notifications
You must be signed in to change notification settings - Fork 714
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Poor quality results for neighborhood queries (zoom level 14) #3538
Comments
The only actionable item here is the part about postcodes. Might be worth pushing them into their own layer. They are currently part of the |
Sorry, but could you please elaborate why the other cases are not "actionable"? Are you saying these are tagging errors or...? |
Lets just take example 1: the documentation says "zoom=13" returns suburb-like object and "zoom=14" returns neighbourhood-like objects. Your complaint: I get a Essentially, this issue seems to complain that zoom=14 is doing what is documented but you would rather that it does something else. This something else is not sufficiently defined to be actionable. On the general difficulty of processing place nodes, I refer to the many issues which discussed the subject. Start with #231 and #575 and work your way through the referenced issues. |
OK, I can understand that it is challenging to improve upon the first case since the areas are not well-defined. But what about the two other cases in Italy? Why is zoom=14 returning coarser-grained results than zoom=13? I assume it is trying to find a place=neighborhood there, but there isn't any. But then why not return the ok-ish result of place=suburb (like the zoom=13 query does)? And, since you are referencing the documentation, why is it that in Trentino we get a place=isolated_dwelling at zoom level 14? OSM Wiki defines those as "the smallest kind of human settlement." Hence that should only ever appear at zoom level 15, "any settlement", nothing above that. |
While queries at zoom level 13 generally work pretty well, I have noticed several inconsistent results for queries at zoom 14. This affects different countries. I have gathered some representative examples here.
Lahti, Finland
https://nominatim.openstreetmap.org/ui/reverse.html?lat=60.95791&lon=25.67280&zoom=14
This matches a place=neighborhood that is almost 2 km away from the original location. The same query at zoom level 13 matches the expected place=suburb (Liipola). It seems that Liipola "loses" just because it is not a place=neighborhood.
Milan, Italy
https://nominatim.openstreetmap.org/ui/reverse.html?lat=45.48304&lon=9.21327&zoom=14
This returns only the admin level 10 boundary (Municipio 3). Meanwhile at zoom level 13 we get a place=suburb (Loreto), which is finer-grained and the expected result. Shouldn't the zoom 14 query give a finer result than the zoom 13 one?
Trentino, Italy
https://nominatim.openstreetmap.org/ui/reverse.html?lat=46.16211&lon=11.10783&zoom=14
This returns a place=isolated_dwelling. According to the docs at zoom level 14 we should be getting "neighborhood", not "any settlement". The isolated dwelling would make sense at zoom level 15. Actually the result to zoom 15 is far worse; it is in a completely different village.
Darmstadt, Germany
https://nominatim.openstreetmap.org/ui/reverse.html?lat=49.84965&lon=8.63038&zoom=14
This returns a postal code. Ditto for zoom 15. The same query at zoom 16 matches the nearest street and contains the expected place=quarter (Heimstätten-Siedlung).
The text was updated successfully, but these errors were encountered: