Replace `desc-plt0p*_mask`s with a log10(p) statistical map?

bbfrederick / rapidtide

rapidtide - a suite of programs for doing time lag correlation analysis on fMRI data

Apache License 2.0

75 stars 14 forks source link

Replace `desc-plt0p*_mask`s with a log10(p) statistical map? #174

Open tsalo opened 1 month ago

tsalo commented 1 month ago

Is your feature request related to a problem? Please describe. I was thinking it might help reduce the number of outputs if you wrote out a log10(p) map of maxcorr significance instead of masks at different thresholds. This isn't something I feel strongly about- I just thought it might streamline things a bit.

bbfrederick commented 1 month ago

That's not a bad idea. I'm not sure if initially I'd want the log10(p) map to be in addition to, or instead of. I mean, it would be redundant, and since you could generate the plt0p* maps from the log10(p) maps, maybe it's not worth doing both.

The nice thing about the log10(p) map is that then you could set the mask p arbitrarily in tidepool...