search

iChaker / CPM_framework

CPM for fMRI voxelwise estimation of arbitrary models
0 stars 1 forks source link

spm_prf_review / spm_prf_response #4

Open SRSteinkamp opened 2 years ago

SRSteinkamp commented 2 years ago

Hi @iChaker,

how much would it break the current implementation, if we would change the get_parameters routine to keep returning the latent parameters?

And implement something like get_real_parameters or something to asses the real parameter values.

If we would use the latent parameters for some function, we could leverage a lot of the PRF toolboxe's utilities, like automatic thresholding based on the covariance etc., confidence intervals and mapping the (latent) parameters + CIs onto the brain.

Best Simon

iChaker commented 2 years ago

It will break BayespRF which will break cpm of course, we can modify this, but I don't see the need. You have direct access to all latent parameters in PRF.Ep without any function calls