[geowave-dev] Determine Tiernumber for Polygons

[Marcel <m.jacob@xxxxxxxxxxx>]

Hello,
I have another theoretical question:
http://accumulosummit.com/program/talks/geowave-geospatial-and-geotemporal-data-storage-and-retrieval-in-accumulo/
Please have a look at slide 18. The blue polygon will be index in tier 
2. This is okay for me.
You say that the red polygon is stored in tier 3. That´s not quite clear 
with the given information from the slide.
Consider the following for red polygon:
tier 0 (all) -> 1 entry
tier 1 (blue) -> 2 entries -> 1 duplicate
tier 2 (red) -> 2 entries -> 1 duplicate
tier 3 (yellow green) -> 4 entries -> 3 duplicates

So when minimizing the number of duplicates it should be stored in tier 
0. If one duplicate would be okay (whatever that reason might be), it 
should be stored in tier 2. But why tier 3? Does it correlate with the 
number of empty cells that a finer-grained tier is used?

Is there another presentation, with some more theoretical details which 
you could provide? Maybe something from internal discussions or 
something like that.

Best regards,
Marcel Jacob.