Allow additional kwargs to pass from reproject_match() -> reproject()

[four43]

I wanted to set a specific nodata value when running a reproject_match and I figured this was the most flexible. What do you all think?

• Tests added
• Fully documented, including docs/history.rst for all changes and docs/rioxarray.rst for new API

[codecov]

Codecov Report

Merging #436 (2cd14e8) into master (1755f90) will not change coverage.
The diff coverage is 100.00%.

@@           Coverage Diff           @@
##           master     #436   +/-   ##
=======================================
  Coverage   96.07%   96.07%           
=======================================
  Files          13       13           
  Lines        1530     1530           
=======================================
  Hits         1470     1470           
  Misses         60       60           

Impacted Files	Coverage Δ
rioxarray/raster_array.py	98.32% <100.00%> (ø)
rioxarray/raster_dataset.py	100.00% <100.00%> (ø)

---

Continue to review full report at Codecov.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update 1755f90...2cd14e8. Read the comment docs.

[snowman2]

This makes sense to me, thanks 👍 . Mind adding a simple test to ensure that the nodata passed into reproject_match works properly?

[four43]

@snowman2 how do you create these fixture files? When trying to re-use the same parameterized fixtures I keep getting this while trying to open mine:

ValueError: Given file dataset contains more than one data variable. Please read with xarray.open_dataset and then select the variable you want.

After verifying in my debugger the file looked correct I just dumped it out to the fixture location with a mds_repr.to_netcdf([fixture_path]) but that doesn't seem to work...

It has the single data variable!

<xarray.Dataset>
Dimensions:                        (x: 150, y: 100)
Coordinates:
  * x                              (x) float64 4.853e+05 4.857e+05 ... 5.429e+05
  * y                              (y) float64 5.029e+06 5.029e+06 ... 4.991e+06
    spatial_ref                    int64 0
Data variables:
    __xarray_dataarray_variable__  (y, x) int16 -9999 -9999 581 ... 867 804 706

[four43]

I squashed and committed what I have but will need help with that fixture. Thanks!

[snowman2]

I am thinking you need to create the netCDF file starting from a DataArray instead of from a Dataset. Are you able to do that?

[four43]

I tried that as well, I gave it the ol:

mds_repr['__xarray_dataarray_variable__'].to_netcdf(modis_reproject_match__passed_nodata["compare"])

And open_dataarray() is still mad! It's pulling in another 'spatial_ref' data_var when opening with that. Did I configure my fixture wrong?

[snowman2]

What happens if you add decode_coords="all" when opening?

[four43]

That did it! It's odd that it was required there for some reason...

[snowman2]

It's odd that it was required there for some reason...

Related to #287

[snowman2]

Thanks @four43 👍

[four43]

Thanks for the review and help @snowman2. We appreciate the project :D