Density estimate through sampling

[e-n-f]

Estimating point density at the highest zoom by sampling the size of the gaps between points rather than using the absolute highest density or a full CDF calculation.

[e-n-f]

Mean of log of density in SF data (in z32 units), when aggregated to different zoom levels:

Standard deviation of log of density in SF data (in z32 units), when aggregated to different zoom levels:

[e-n-f]

Actually the first problem with this is that the sampled density calculation comes out different based on how the data was bucketed during import, so I need to figure out how to fix that before trying to derive anything else from it.

[e-n-f]

For my future reference, before I forget again, points that are quadkey distance N apart are approximately sqrt(N)/33 feet apart.