Bake in county outlines

[jrleeman]

Cartopy doesn't have state outlines - adding them is too difficult. We should bake this in somewhere.

[ktyle]

I would also advocate baking in US County outlines. I don't see them in NaturalEarth.

[dopplershift]

We'll see about getting states in CartoPy first SciTools/cartopy#888

One of the challenges is finding a properly licensed and reliable counties dataset.

[dopplershift]

Ok so SciTools/cartopy#888 is merged, so we'll have states in the next release of CartoPy.

Now...counties.

[kgoebber]

I've started work on generating a counties shapefile with the scale of 50m based on the Natural Earth 50m states, provinces, and lakes shapefile. It'll take some time, but I've got a process that I think will slowly, but surely get us there. The original county shapefile is from U.S. Census Bureau, should be okay licence-wise.

[kgoebber]

Okay...so the method I have to do this would take a long LONG time to complete. The primary difficulty lies in the county polygons from the U.S. Census Bureau are available at 500K, 5M, and 20M resolutions (none of which align with the Natural Earth data [10M, 50M, 100M]. So what I have done (with the help of one of my geography colleagues) is to create new state boundaries based on the county boundaries using QGIS and the Dissolve method. I now have coordinated county and state lines for 500K, 5M, and 20M resolutions. Having these shapefiles as a part of MetPy might be the best short (maybe even) long term solution to this problem.

I should add that it is not "clean" to get these shapefiles in with Cartopy, but it can be done. Maybe we bake in a simpler format for getting these files read in and able to be used in making a map.

http://nbviewer.jupyter.org/url/fujita.valpo.edu/python/counties_test.ipynb

[jrleeman]

What's the size of the shapefiles?

[kgoebber]

The 500K county file is 16.8mb, state outlines 4.7mb.

The 5M county file is 3.6mb, state outlines 1.4mb.

The 20M county file is 1.1mb, state outlines 225kb.

[kgoebber]

The 500K file loads pretty slow, but the 5M and 20M are pretty quick.

[jrleeman]

I'm thinking these need to get pulled from a remote and cached. Too much to put in the individual releases. This would be similar to what Cartopy does with the natural earth datasets. Question is does it go upstream or here?

[kgoebber]

Shapefiles are now available at http://fujita.valpo.edu/python

[jrleeman]

@kgoebber - do you have the associated DBF files to go with these? Otherwise, can't plot.

[kgoebber]

All needed files now located at http://bergeron.valpo.edu/python/states_counties/

…

-- Kevin Goebbert, Ph.D. Associate Professor of Meteorology Kallay-Christopher Hall, 201-F 1809 Chapel Drive Valparaiso, IN 46383 (219) 464-5517 <//(219)%20464-5517> fax: (219) 548-7738 <//(219)%20548-7738> https://kevingoebbert.youcanbook.me/ On August 21, 2018 at 11:36:50 AM, John Leeman ([email protected]) wrote: @kgoebber <https://github.com/kgoebber> - do you have the associated DBF files to go with these? Otherwise, can't plot. — You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub <#447 (comment)>, or mute the thread <https://github.com/notifications/unsubscribe-auth/AT4Y1e4aRgQkvwhGFXW_wrmZo-CZ0-AWks5uTDcigaJpZM4Nxhlu> .