Expose testing methods

[spencerahill]

Similar to numpy's numpy.testing module, I would find it useful for xarray to expose some of its testing methods, particularly those in its base TestCase.

I imagine other folks whose packages and/or modules derive heavily from xarray would find this useful too. For now I am just copy-pasting xarray's code for these methods into my testing modules.

[jhamman]

👍

[shoyer]

Most of these are as simple as assert d1.identical(d2) or assert d1.equals(d2) (see here for more details on these methods). I'd be OK with adding these methods in a dedicated testing module, though.

[spencerahill]

Indeed. I'm happy to do this sometime (eventually...), but if somebody ends up with a pressing need for this in the meantime, go for it. Thanks!

[jhamman]

@spencerahill - have you made any progress here? If not we (@dgergel) may be able to help out.

[spencerahill]

Thanks for the ping. Sorry, I haven't. I'm actually submitting my PhD thesis in ~2 months and then defending it this summer, so I won't be able to help much until August or September. So by all means, go for it.

[fmaussion]

I'd be willing to prepare a PR, if this is still actual. Today I needed the *AllClose methods for my tests, I think it would be great to have these methods exposed.

I think I have an idea on how to proceed, but maybe I am underestimating the required amount of work.

Numpy for example has the np.allclose and np.testing.assert_allclose functions. Should xarray also have a xr.allclose function or are we happy with a xr.testing.assert_allclose alone? Also, the current xarray implementation has a different function for variables, datasets and coordinates: could these be merged in a generic assert_allclose which takes decisions on what to do based on the input type?