search

plexo1 / ratrix

Automatically exported from code.google.com/p/ratrix
0 stars 0 forks source link

spatial analysis #13

Open GoogleCodeExporter opened 8 years ago

GoogleCodeExporter commented 8 years ago 
what we really want is a motor readout of confidence.  a fast direct response 
is certain, a correction from one direction to the other is not.  this isn't 
well-separated from diligence -- noisy meandering tracks may not be low 
confidence, just satiated.

secondarily, we want to see changes over time (long-term motor learning or 
short-term satiation/fatigue within session)

to parameterize tracks, how about fitting n-order polynomials to them and 
pca'ing the coefficients?

Original issue reported on code.google.com by erik.fli...@gmail.com on 7 Jun 2012 at 5:52

GoogleCodeExporter commented 8 years ago 
chunk track figures by session (where session defined by grey lines in figure 
-- not ratrix start/stops) -- show zoomed in enough w/horiz scroll to see ~200 
trials per screen.

Original comment by erik.fli...@gmail.com on 7 Jun 2012 at 6:37