Skip to content

Instantly share code, notes, and snippets.

@talllguy
Last active August 16, 2018 18:04
Show Gist options
  • Star 0 You must be signed in to star a gist
  • Fork 0 You must be signed in to fork a gist
  • Save talllguy/7d813ece238f359317786a18f7b7bbcb to your computer and use it in GitHub Desktop.
Save talllguy/7d813ece238f359317786a18f7b7bbcb to your computer and use it in GitHub Desktop.

On 2016-08-26 22:06:07 UTC annapolissailor wrote:

Hey there (what can I call you?) :)

Hey Elliott, I'm Richard! Just wanted to send a quick email saying that I haven't forgotten about this. This past week we had a sailing camp come through which kept me pretty busy.

Sounds like a good plan regarding simplification. While you're de-noding, I suggest adding some more address information, like the city, state, zip, all of which are in that file. Here's an example of some processing I did for the Baltimore County Import. If I were you, I would email that list after you come up with a plan to fix the simplification. The community can be pretty finicky, and so to avoid having someone go and revert all of your hard work, it is best to reach out. I can help with that as an ambassador of sorts, if you'd like.

I have no problem with adding city/state but it seems rather redundant as all the major search engines from what I can tell use run time point in polygon no? Definitely have no problem adding this if it is the usual thing to do however. I'm a bit more hesitant to add zip. USPS changes these on a near yearly basis without much warning. From what I could find on the wiki, Nominatim uses census data to calculate these.

Also in that GitHub repo for Baltimore Co, you'll find some scripts for address expansion, which you could adapt for this import, so you don't need to go and try to manually expand the address points in the file.

I'll take a look! I haven't minded the manual stuff too much as a lot of the streets are terribly messy and it forces me to review each one manually which should lead to higher quality data for both the addresses and the street network.

Finally, I've reviewed this with some friends, and some have balked at the building=yes for everything. In the address file, there is a BLDG_CODE field. You could use that attribute to set the building type on OSM. For instance, RESIDENTIAL in the data would equal building=residential.

Good idea! I'll take a look! I should be able to generate centroids for all the buildings and merge this attribute.

There are some great QA tools out there to find missing roads based on the address info. That is one of the great things about importing addresses; you'll find street issues as well. Here's my favorite one, showing an issue you've identified by importing addresses along Island View Drive, which is Island View Road on OSM!

This is absolutely fantastic. No idea this existed.

I'm going to be pretty busy this weekend but next week I'll draft up an email for the import group you mentioned and run it by you.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment