search

mobeets / nullSpaceControl

1 stars 0 forks source link

assess hypotheses in spike space #220

Closed mobeets closed 8 years ago

mobeets commented 8 years ago

could assess this in spike space easily enough, but estimates might be much noisier (e.g., estimating covariance in high-d with low numbers of points)

mobeets commented 8 years ago

spike scores agree basically perfectly in mean error at least. not in coverror.

note how 20120709 is the only one where hab is lower, just like with latents i corrected predictions to be 0 if below 0.

screen shot 2016-07-12 at 12 07 56 pm screen shot 2016-07-12 at 12 08 01 pm screen shot 2016-07-12 at 12 08 07 pm screen shot 2016-07-12 at 12 08 14 pm screen shot 2016-07-12 at 12 08 17 pm

mobeets commented 8 years ago

but as expected, cloud-og (cloud-200) fixes the covariance problem:

screen shot 2016-07-12 at 12 27 10 pm screen shot 2016-07-12 at 12 27 16 pm screen shot 2016-07-12 at 12 27 20 pm screen shot 2016-07-12 at 12 27 25 pm screen shot 2016-07-12 at 12 27 30 pm

mobeets commented 8 years ago

p.s. in those last plots, the bottom-left panel should have the ylabel "covError"