Functions for stack mangement

[Bartesto]

Possibly take mosaicked by KM annual data:

• Mask and clip to standard extents
• ability to add in new annual by either recreating stack for each annual monitoring update or add to a previously created stack??

[Bartesto]

Good progress on:

• ndvi ingest
• use of stars proxy objects to delay calcs
• processing > crop > to pfc > sensible data range (10-100)
• out of proxy to raster for > invert masking clean up > write to disk

[Bartesto]

Now have:

• veg_dens() which calculates veg density from coeffs and applies to any raster/s
• mask_product() which can apply any number of masks to raster/s

[Bartesto]

re-wrote and tested all functions up to and including creating veg classifications

[Bartesto]

Now written a parallel version of the veg_dens() workflow. Cuts processing time in half

[Bartesto]

Now have issue with masking "in" a value for cloud. Looking at various ways:

• currently a rasterized colud mask can "cover" and replace NA values in target raster but gives value to water

[Bartesto]

Worked back on the trendclass function to double check outputs. Have come to conclusion that we can just use the slope from the regression to classify trends.

Also have a working example of doing freq type table calcs for whatever you point it to for reporting regions.

[Bartesto]

Worked most of the week on integrating clouds from cloud masks and lastly a "cloudy probabilities" veg classification. Have to write the easier non cloudy version next and combine both into a wrapper function.