I am new OSM, and I like it very much. I am analyzing a set of addresses in the US states of Rhode Island and Massachusetts, primarily needing geocoding from address to lat/long. To conserve resources, I downloaded those state’s data from geofabrik and combined them with osmium-tools and then built Nominatim appropriately. Things seemed to be going ok, until I realized many addresses in Rhode Island were not found and/or were being matched to similar addresses in Massachusetts. I did some tests, and for the addresses that were not found, if I removed the state, it would frequently find the address. After I did some digging, I think I narrowed it down to the fact that there is not a level 4 administrative boundary for Rhode Island in my combined file. There was a level 4 administrative boundary for Massachusetts. Is there supposed to be a level 4 administrative boundary for Rhode Island in the Rhode Island specific file, or am I doing something wrong?
I suspect what is happening is the RI boundary might be getting clipped on the geofabrik side, but they would have to confirm or deny that.
Yes, the RI clipping boundary was missing a tiny bit on the water. I have fixed it now and the next extract (due in a couple hours) should have the complete boundary relation. The clipping polygons are now kept on GitHub GitHub - geofabrik/polygons_download.geofabrik.de: clipping polygons used at https://download.geofabrik.de/ to make it easier to report issues.
2 Likes