search

GordonLudlow / s2geometryserver

simple http server built on the s2geometry library
0 stars 0 forks source link

Thanks! #1

Open ryantheleach opened 2 years ago

ryantheleach commented 2 years ago

Sorry, I didn't have a better way to contact you, thanks for making this mini server.

I don't know the level at which the cells are at, but I have it on pretty good authority from someone I trust to have done their research (years and years ago), that the Australian MU data source was based on LGA's and a previous census.

Unfortunately I can't remember the year, or what resolution they claimed it was.

I would suggest looking for LGA areas that have had recent developments, measuring the MU density in those areas, to try to determine which census year the data is from.

From there, depending on the accuracy of the data available (remember, this initial data would have been from google, not niantic before they spun out) it would be interesting to take that data and average it across different level s2 cells, and attempt to see what S2 cell level seems to have the best correlation between small fields. on either side of the boundaries.

Also a similar project has been undertaken previously. Ask around and make noise in global channels about Bovine Ethology.

GordonLudlow commented 2 years ago

Hi Ryan, I haven't worked on this in a bit, but while I was at it I made a dev log to keep track of the experiments. https://gordonludlow.com/ingress/reverse-engineering-mu/ I'm pretty sure it's level 13 S2 cells, at least in the Seattle area.

"Bovine Ethology" is a hilarious name!

ryantheleach commented 2 years ago

Looking back at old data, and going by memory, 13 seems a little too detailed for our local area.

It's my belief upon review, that it's more than likely 12 or 11.

I can share a copy of this, but for privacy reasons (contains linked google accounts, emails, names of a few other members (it's a google doc)) I don't want to share the link publically. Do you use discord at all?