Wind rose

[jrleeman]

It could be useful to add a wind rose plot. I think we could just use polar bar plot?

[dopplershift]

Well, this exists: https://github.com/scls19fr/windrose

[jrleeman]

This came up on stackoverflow today (https://stackoverflow.com/questions/49419349/making-a-windrose-in-python-from-degree-wind-direction-and-speed). @akrherz pointed out that the windrose package is not compatible with modern matplotlib and his fork has some improvements. I'd say given the number of windrose users out there, this should bump up slightly on the list.

[dopplershift]

@akrherz have you tried submitting your improvements to: https://github.com/python-windrose/windrose ?

[akrherz]

Hi @dopplershift , I honestly don't remember and unsure why I would not have attempted PRs with these. shrug.

[dopplershift]

I'm not against including the functionality in MetPy, just wondering if the better thing is to play ball with the other project (which has taken the old code and moved it forward).

[jrleeman]

After about +4 at workshops, we need to move this up on the list.

[ocefpaf]

Just found this issue and, as someone who took up developing windrose recently, I can say:

1. it is kind of up-to-date with latest mpl for a while and most bugs are being addressed;
2. I would be OK moving it to metpy if there is interest on both sides (I don't speak for all windrose devs but my guess is that they'll be onboard too).

[dopplershift]

@ocefpaf I'd say we'd be happy to incorporate it if there's interest on the windrose end.

[dopplershift]

@ocefpaf Are there any tests? 😉