search

gkaguirrelab / rgcDisplacementMap

tools to model retinal ganglion cell displacement from receptive field locations in the human retina
1 stars 0 forks source link

Try using proportion cumulative cone density for the mRF:cone linking function #8

Closed gkaguirre closed 7 years ago

gkaguirre commented 7 years ago

Current done in terms of max cone density. Perhaps proportion of the cumulative density (as is done for the mRGC:RGC function) would work better across meridians.

gkaguirre commented 7 years ago

Explored this and found it doesn't do so well. Sticking with the existing approach based upon proportion max cone density.

I note that the proportion max cone density regularization brings the Watson 2014 prediction for the cardinal meridians into excellent alignment, with the exception of the superior temporal meridian.