try re-running with no variance regularization

davidwhogg / TheTractor

contributions to and ideas about the Tractor (which mainly lives at Astrometry.net)

http://theTractor.org/

10 stars 3 forks source link

try re-running with no variance regularization #24

Closed davidwhogg closed 12 years ago

davidwhogg commented 12 years ago

I think this regularization term is useless; let's drop it and re-run, maybe WHILE THE PAPER IS IN REFEREEING?

davidwhogg commented 12 years ago

this doesn't work -- we do need a tiny bit of regularization. But I lightened it considerably with no changes, so that's good. I also make it better defined (long story).