This is somewhat a complement issue to the treatment of xarray non-dimensioned coordinates #5214 .
That issue concerned the behaviour non-dimensioned with auto-populated widgets. With good reason non-dimensioned coordinates are not treated as kdims in that case.

However I think it would still be very useful if they could be used with kdims when explicitely specified by a user. Consider the following example from the xarray docs:

In [65]: ds.coords
Out[65]: 
Coordinates:
    lat             (x, y) float64 42.25 42.21 42.63 42.59
    lon             (x, y) float64 -99.83 -99.32 -99.79 -99.23
  * time            (time) datetime64[ns] 2014-09-06 2014-09-07 2014-09-08
    reference_time  datetime64[ns] 2014-09-05
    day             (time) int64 6 7 8

Here 'lat' and 'lon' are non-dimensioned coordinates, yet I would think that it makes sense to want to use them as kdims to a scatter plot or heatmap.

As another example, consider a simulation that has multiple variants of its time axis:

• simulator step (int)
• simulation time (float)
• real-world time (float)

Only one of this can be a dimensioned coordinate in the DataArray, yet all of them could sensibly be used as a plotting axis.

Finally, it may be worth noting that plotting is one of the explicit intentions for non-dimensioned coordinates according to the docs I linked above.

Describe the solution you'd like

Given an xarray DataArray,

da = xr.DataArray(np.arange(6).reshape(6,1), dims=["x", "stat"], coords={"x": np.arange(6), "stat": ["mean"], "x2": ('x', np.arange(6)/6)})

it is very easy to plot using dimensioned coordinates as kdims:

hv.Curve(da, kdims="x", vdims="stat")

I would like the same to be true for non-dimensioned coordinates

hv.Curve(da, kdims="x2", vdims="stat")

Workaround

One approach that already works today is to first convert to a dataframe

hv.Curve(da.to_dataframe(name="mean"), kdims="x2", vdims="mean")

This isn’t too bad (even though the vdim name needs to be repeated), but it still feels like xarray’s annotation abilities (here to specify alternative axes) are being wasted.