Improve on elm/pipeline/predict.py and elm/pipeline/serialize.py. Currently the tests only test that the pickles or netcdf outputs show up. One improvement may be GeoTIFF output or a preparation for datashader.
This is currently handled by passing all metadata from final X sample into the prediction Y ElmStore which can be used for plotting, using the tools of xarray.DataArray.plot
Improve on
elm/pipeline/predict.py
andelm/pipeline/serialize.py
. Currently the tests only test that the pickles or netcdf outputs show up. One improvement may be GeoTIFF output or a preparation for datashader.