Tony McGovern

Thanks for your work @almccon. A few questions and comments: ### `2016_US_County_Level_Presidential_Results.csv` Where do the results for Alaska come from? Did you re-run any of the notebooks on this repo...

@almccon I merged #21 to the master branch, which solves the ward- and district-level results for DC and Alaska, respectively. For DC results, I aggregated all ward-level results to the...

@iraykhel The notebook that compiles the 2016 county-level election results pulls together estimates published major news organizations. It's possible that the New York Times updated their site with actual raw...

Agreed. The issue was fixed in "US_County_Level_Presidential_Results_12-16.ipynb" notebook and still only exists in the "2016_US_County_Level_Presidential_Results.ipynb". I'd like to delete that last file, but don't know whether it's used by anyone....

@d-alex-hughes Forgive me for the delayed response. If you're still able to do that, then we can take a look at the results and accept your PR.

Those are good catches. I did not factor in error catches and should probably do so. Some counties publish incorrect data or miscalculate totals. I'm thinking of creating an output...

@seesharp15 Forgive my delayed response. It looks like you changed the files to factor in Shannon County's (Oglala Lakota) name and FIPS change. You also made updates to county-level totals...

That sounds like the right approach. Generally, Alaska election results are reported at the state-level. See here as an example: [Alaska](http://www.politico.com/2016-election/results/map/president/alaska/) Compare to this: [Arizona](http://www.politico.com/2016-election/results/map/president/arizona/) My map has Alaska with...

@stoneyv openelelctions is also working on translating boroughs -> precincts. It's a difficult task. Here's their project page (nothing on Alaska yet): https://github.com/openelections/openelections-core

This is a difficult task, one without an obvious solution. I'm inclined to remove Alaska from the overall results so national turnout numbers align with county-level components. An approach I've...