Fix x ranges upon projection

[ahuang11]

Data no longer disappears upon zoom as described in #673 (comment).

Closes #673

Screen.Recording.2024-04-12.at.10.01.04.AM.mov

import xarray as xr
import panel as pn
import holoviews as hv
from holoviews.operation.datashader import rasterize
import geoviews as gv
import cartopy.crs as ccrs

hv.renderer("bokeh").webgl = False
import hvplot.xarray
ds = xr.tutorial.open_dataset("air_temperature").isel(time=0)

rasterize((gv.Image(ds, ["lon", "lat"], ["air"]))) * gv.feature.coastline()

[jbednar]

Thanks for addressing this. I can't quite decide if it's the right fix, though; would be good to have e.g. @philippjfr think about it.

[ahuang11]

My other three thoughts about approaches to fixing this is:

1. require -180 to 180 for PlateCarree, else raise an error
2. same as 1, but raise a warning instead
3. if 0-360 is detected, automatically convert the data (potentially costly to load data)

This PR's fix only updates two integers (the x_range), which is basically "free"

I don't think this can be patched in HoloViews either because the logic here https://github.com/holoviz/holoviews/blob/main/holoviews/operation/resample.py#L139-L140 is correct although I don't know what the difference between x0 and ex0

[TheoMathurin]

Maybe this can possibly also fix #593?

[ahuang11]

Maybe! If it uses lat/lon, then there's a chance.

[ahuang11]

The changes work when it's not overlaid with coastline.

import xarray as xr
import panel as pn
import holoviews as hv
from holoviews.operation.datashader import rasterize
import geoviews as gv
import cartopy.crs as ccrs

hv.extension("bokeh")
import hvplot.xarray
ds = xr.tutorial.open_dataset("air_temperature").isel(time=0)

rasterize(gv.Image(ds, ["lon", "lat"], ["air"]))

However, once coastline is overlaid, it produces a blank image.

[philippjfr]

I think this is too eager, specifically it should check that it's actually in the 0-360 range rather than merely all being east of Greenwich, so something like:

          self._unwrap_lons = 0 <= x1 <= 360 and 180 <= x2 <= 360

[philippjfr]

This should be:

if plot._unwrap_lons and (-180 <= x0 < 0 or -180 <= x1 < 0):