Ryuichi Arafune

Thank you for your comment. I'm glad to hear that the implementation process won't be too difficult. I look forward to seeing it implemented soon.

hv.Image((cut2_band.coords["kp"].values, cut2_band.coords["eV"], cut2_band.values)) The above code works well. So I believe the problem should lie in the process for determining the gridded.

hv.QuadMesh has same problem. Further, While `dataarray_a.plot()` and `daarray_b.plot()` show the same ouptuts, ```python hv.extension("matplotlib") hv.Image(dataarray_a) hv.Image(dataarray_b) ``` shows different results.

> Thanks for pointing out this behavior. In your opinion, what should be the determining factor about what goes on the horizontal axis and what goes on the vertical? Should...

I feel that the problem would be that there is no policy for xarrays to plot by using holoviews. I think it is important to determine what that policy is....

@jbednar I understand your point of view. I'm just a very novice user of holoviews, and I didn't know the history of this library. Thus, it might be natural to...

Triple check might also be important. However, I believe that it is better to first try to fix the problem and then consider what to do about it than to...

@ahuang11 I also feel there is no "elegant" solution. However, there is a solution for keeping the holoviews robust. -- (as I said) to make it impossible to plot Datasets...

Here, I would like to point out that both ``dataarray_a.plot()`` and ``dataarray_b.plot()`` is same. ----- Suppose that: ```python import numpy as np import holoviews as hv import xarray as xr...

I said: > To begin with, it is really ambiguous what hv.Image(ds_abc) should show. Suppose that ```python ds2 = xr.Dataset({"B": dataarray_b, "A": dataarray_a}) ``` This is "essentially" identical dataset object...