I have realized I only replied to Clifford and not replied all, so our conversation has not been uploaded to talk-us. Here is the summary of the convo. ________________________________ On Thu, Jun 7, 2018 at 9:53 AM Mulea, Gregory <gregory.mu...@ppd.mncppc.org<mailto:gregory.mu...@ppd.mncppc.org>> wrote Thanks for the reply. I have updated the wiki to include information about our tagging plans. Let me know if that is sufficient. Honestly I’m pretty new to OSM (I have done small basic edits but nothing compared to this) so I’s still learning. I plan on using the JOSM editor to import these files but if there are better options I am happy to learn about them. As for the conflating issue, I believe that JOSM has a method to check for overlap of polygons in which you can decided whether to keep or discard the copy (I plan to do this with the buildings polygon file). As for point data, I am unsure as to how to test/check for already existing points (for example importing the libraries points file, if there is already an existing library I have not found a way to check for that). ________________________________ Breaking the import up into small sections and using the Tasking Manager to work each section at a time is what I've done in the past. I use census voting districts - they seem to be a reasonable size. I believe you can use a python script or like I do, use postgresql to grab polygons/points by voting district. Ask Ian on Slack for permission to administer his TM. JOSM has a conflation tool, but it was broke last time I tried to use it. If there is an existing building outline, I determine which one more accurately represents what's on the ground. If the import is, I select both outlines and use the "Replace Geometry" tools that comes with the utils2 plugin. If the OSM data is better, I copy any tags missing on the OSM data from the import over to the existing building.
Check out https://wiki.openstreetmap.org/wiki/Kirkland_Import<https://na01.safelinks.protection.outlook.com/?url=https%3A%2F%2Fwiki.openstreetmap.org%2Fwiki%2FKirkland_Import&data=02%7C01%7Cgregory.mulea%40ppd.mncppc.org%7C868eaf91971a4477684c08d5cc99dc99%7Ca9061e0c24ca4c1cbeff039bb8c05816%7C0%7C1%7C636639883579760784&sdata=i5BfSY2cDl8F%2BYEBsY6%2BZAsPRbiSzgmoQ5iWaVH%2BHI4%3D&reserved=0> for information on how we did that import. Best, Clifford ________________________________ On Thu, Jun 7, 2018 at 10:51 AM Mulea, Gregory <gregory.mu...@ppd.mncppc.org<mailto:gregory.mu...@ppd.mncppc.org>> wrote: Breaking up the import seems like a good idea. I use ESRI’s ArcMap, how do the scripts compare between ArcMap and what you used? Also, I have a relatively small file of 43 library points. Would it be possible to do somewhat of a test import in order to test JOSM’s conflation tool and figure out different workarounds? Thanks ________________________________ The 43 points seems like it would be perfect to do with one import. I use org2osm.py to translate shapefiles to .osm files. With just a bit of knowledge you should be able to write the translation script to go from shapefile tag to .osm tags. Then merge your library point with the existing. I'd be curious to learn how the conflation tool works. Instead of using ogr2osm.py you could load the shapefile into JOSM and fix the tags by selecting all of points to change tags. For example, change NAME=<a whole bunch of names> to name=<a while bunch of names>. Depending on your address information, that make take more work. We use addr:housenumber, addr:street (with a spelled out street, no abbreviations), addr:city and addr:postal_code. QGIS is the only application I've ever worked with. Most of what I do is run from the command line using postgis and gdal tools with the help of the ogr2osm.py translation tool. ________________________________ Great, would I be able to do the test import today? I also have a small file of buildings to test with polygons too. I plan on just using the JOSM OpenData Plugin to load in the files and then changing the tags through there but I could give the py script a shot. Is it necessary to split the addr tags into housenumber and street or would it be okay to leave it as addr:full? ________________________________ Go ahead with the test import. When you are done, can you point to the area for review? ________________________________ Great, The area will be in Prince George’s County, Maryland
_______________________________________________ Talk-us mailing list Talk-us@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk-us
