Greg Lucas

The vectors are assumed to be in the "projection" coordinates, not necessarily the geographic coordinates. See this PR for an attempt to rectify some of this: https://github.com/SciTools/cartopy/pull/1926 which I think...

I think this is because `contourf` is calling autoscale to readjust the limits after you've set the extent. What happens if you move the `ax.set_extent()` call just before the show...

I'm guessing we'd need to add a GeoQuiver and GeoBarb collection to handle these updates properly, similar to what we've done for GeoQuadMesh: https://github.com/SciTools/cartopy/blob/main/lib/cartopy/mpl/geocollection.py which we then return in `pcolormesh()`...

10m oceans are known to be extremely slow, so I would definitely not recommend using those if you want speed. See this thread for some performance suggestions: https://github.com/SciTools/cartopy/issues/1895 Unfortunately, no...

My guess would be that the wrapped cells are different between the cases and there are some different autoscaling limits associated with those cells. https://github.com/SciTools/cartopy/blob/a8f62baf12e73921a5910a61f2c7b75d8dfc07fb/lib/cartopy/mpl/geoaxes.py#L1956-L1990 When we call `self.pcolor()` there...

Having some kind of cache for updating image data without needing to transform every time seems like a good improvement! But in your example, you are then putting that into...

I agree you'll need to do some kind of refactor to cache the needed quantities for re-use, so that may be a good option.

Perhaps lexsort applied to x and then y would be useful here instead of just argsorting the x values? https://numpy.org/doc/stable/reference/generated/numpy.lexsort.html However, the key difference between transform first and the default...

If you reverse the order of the coordinates, that should fill the internal side. You might also want to create a patch instead which I don't think would have that...

You shouldn't need to know the web address for the high-resolution states/coastlines. The tricky part is realizing that Cartopy can get these for you and what resolutions are even available...