jmp75 / silverpieces Goto Github PK

Integrate https://github.com/CSIRO-enviro-informatics/c3dis-2019-pangeo/blob/master/All_Catchment_Polygon_Mask_Averages.ipynb and (preferably) salem base approaches to cookie cutting / producing masks.

I suggest the first iteration on this might be to add new functions to functions.py that accept geojson or files, a target xarray template (or maybe another kind of description of spatial dimensions) and generates a mask xarray. Additional functions might save the mask to disk and load and apply the mask to real datasets.

Measure runtimes from the '02' notebook using the Tasmania daily subset. Extrapolate to continental scale.
Presence of missing values (sea) may well prevent extrapolation of runtime to the continental scale, however. semi-synthetic data may be a good idea.

On the back of issue #4, as we except a need to parallelise operations, what are the options with dask or other tech?

This is a test of the installation instructions to make sure nothing is unduly tied to our organisation or assumed.

Fill out instructions for use of conda.

Also a test issue to use a project board

testing edit - ben