Hardening / documenting whitebox_tools xarray wrapper + related CI

[PeterDSteinberg]

Hi @dharhas Here are some estimates for the Whitebox tools work required to finish the xarray wrapping. I think this issue list captures everything. Let us know if you think this is reasonable ( cc @jbednar @gbrener ).

• 14 hours - Deduplicate code / cleanup in whiteboxtools/whitebox*.py #15
• 12+ hours - Docstrings / doctests / documentation in ./whitebox_tools ./whitebox_tools/tests #14
  • Note that the third bullet on the issue 14 above is an extended notebook example in Elm related to Lidar and the first two bullets are more self-contained and specific to Whitebox Tools. The first two bullets on 14 are what I think is 12-ish hours, ideally for ERDC money. There are several options for the third bullet extended Elm Lidar notebook's time, or a combination of them, we may consider: 1) do the extended Lidar example with ERDC funds on ERDC schedule, 2) do the Lidar example with NASA funds but the work would need to be deferred.
• 4 hours - Ensure numpy and xarray are optional dependencies #13
• 16 hours - Tools unable to use xarray opened with xarray.open_rasterio even though the file can be used directly #12
• 24 hours - Build with conda-forge #4

[PeterDSteinberg]

Hi @dharhas , One more issue that is not in the list above:

17 - Do we want to support Laz/Las Lidar data files now and/or images that are inputs to some Whitebox Tools? My thought is that supporting the two file types would be about 20 hours of work. Perhaps we can approach that tentatively by starting with half day of testing first to see exactly the failures and then touching base with you?

cc @jbednar @gbrener

[dharhas]

@kajiglet @sdc50 please have a look and weigh in. In my mind, we need to split this into two pieces, 1) the minimum parts that need to be done to get it merged back into upstream and built on conda-forge so things are sustainable and 2) Things that would be nice to have and could be done later either if we have funds left or if we have developer time at ERDC we can contribute or if its already in the Nasa timeline somewhere.