Ryan May
Ryan May
So it looks like we use the `savefig.dpi` rcParam if no DPI is passed to `save()` on the animations, and `to_jshtml()` does not let you set it. The thing is,...
What happens if you set `rcParams['savefig.dpi']` to a higher value?
Just to follow on: given that many (most?) calculations deal with entire numpy arrays of values, catching these kinds of regions of validity can be a bit more tricky. Assumptions...
I'll just note that an alternative is to set default permissions on the whole repo in the settings:
As far as I can tell, this is merely an oversight on our part, because we usually intend for plots to do that (as long as it's sensible). Would be...
I think just returning directly the object created (and happens to be stored as an attribute on the object) is sensible and consistent with the rest of the class, as...
Strictly speaking, `vector_derivative` assumes (for purposes of math with x/y-based vectors on a Cartesian plane) that u and v are vector components. So I don't think this implementation will work...
Looks like GEMPAK [assumes scalars](https://met.nps.edu/~ldm/gempak/apxB2.pdf), so we can do that too. Probably should be using `geospatial_gradient()` like `advection()` does--which matches the `check_units()` you wrote above too.
> I really don't think we should be marking packages we ship with bounds like `>=3.6` when we have not shipped python 3.6 in a long, long time and no...
> But we do still ship 3.6? I just created a perfectly fine 3.6 environment on my mac. I would be extremely confused if I went to install an upstream...