Incorrect results from polyfill about h3 HOT 8 CLOSED

Can you publish the compiler/operating system/CPU architecture you are running on?

from h3.

Sure.
I'm using Suse Tumbleweed, Linux 4.17.12-1-default #1 SMP PREEMPT Fri Aug 3 06:36:01 UTC 2018 (ef4920c) x86_64 x86_64 x86_64 GNU/Linux
make is GNU Make 4.2.1 Built for x86_64-suse-linux-gnu
gcc is gcc (SUSE Linux) 8.1.1 20180719 [gcc-8-branch revision 262874]

from h3.

I was able to reproduce this issue using your Java code. I got 349 hexagon IDs, and I see they match some of the addresses you got.

I believe the problem here is that H3 expects coordinates in the latitude,longitude order, but GeoJSON uses longitude,latitude order. When I swap the coordinates, I get 282 hexagon IDs and they match @nrabinowitz' set.

from h3.

Wow good catch.

from h3.

Ah. I'm so glad you recognized that. I had them the correct way in an earlier attempt, but the output was transposed. Everything works perfectly when I deal with it on the output, too.

Thank you. I'll close this.

from h3.

Have this issue been fixed? In a GeoJSON the coordinates are in the order [lon,lat]. However, the polyfill function expects the coordinates to be [lat,lon]. So we aren't truly following the GeoJSON standards.

from h3.

Just saw the issue #141 ! Will move to the python bindings repo! :)

from h3.

Have this issue been fixed? In a GeoJSON the coordinates are in the order [lon,lat]. However, the polyfill function expects the coordinates to be [lat,lon]. So we aren't truly following the GeoJSON standards.

The core library doesn't attempt to follow GeoJSON - this is left as a job for the bindings, and in some cases additional libraries, to handle.

from h3.