Plotting an xarray of size (nx, ny) in 2D whatever nx and ny

[patquem]

Hello,

I work with 2D arrays of variable sizes and I wish to represent them (always) in 2D representation.
Unfortunately, when one of the sizes is set to 1, xarray.plot() automatically degenerates the plotting toward a 1D signal representation.
Is there a way to prevent this ? (like an hidden option to force the 2D representation in any circunstances ?).
Thanks,
Patrick

import numpy as np
import xarray as xr
import matplotlib.pyplot as plt

nx = 2 # OK
# nx = 1 # NOK -> 1D signal
arr = np.random.random((nx, 10))
x = np.arange(nx)
y = np.arange(10)
xarr = xr.DataArray(arr, coords=[x, y])
xarr.plot()
plt.show()

[dcherian]

See #620

[patquem]

Thank you dcherian for your (very fast) answer
Indeed the post #620 is similar to my request. But I don't see what it has been done finally to respect the request ?
I can't believe it 's still opened since 2015 ? @jhamman ?

[johnomotani]

A simple workaround might be to call imshow explicitly, like

import numpy as np
import xarray as xr
import matplotlib.pyplot as plt

nx = 1
arr = np.random.random((nx, 10))
x = np.arange(nx)
y = np.arange(10)
xarr = xr.DataArray(arr, coords=[x, y])
xarr.plot.imshow()
plt.show()

pcolormesh() instead of imshow() doesn't work for me - shows a blank plot even with infer_intervals=False, not sure why...