Thanks! I'll try that.
Mihaela
On Fri, May 29, 2015 at 12:20 PM, Douglas Greve gr...@nmr.mgh.harvard.edu
wrote:
The clusterwise p-value (--cwp) will correspond to a certain cluster
size, so you can play with that to get clusters of a certain size. In the
vbm example you give below, each of
The clusterwise p-value (--cwp) will correspond to a certain cluster
size, so you can play with that to get clusters of a certain size. In
the vbm example you give below, each of those clusters will have a
p-value, they just choose not to report them.
doug
On 5/29/15 11:50 AM, Mihaela Stefan
Hi Doug,
The simulation went smoothly, however, as I suspected, no clusters were
detected. I wonder if you have any thoughts about my question in the
previous email, regarding the uncorrected p values.
Thanks!
Mihaela
On Thu, May 28, 2015 at 11:09 AM, Mihaela Stefan mikaelastep...@gmail.com
Thanks, Doug, the command seems to work. I guess it will take a while to
finish running.
On a different note, I have a large data set that doesn't seem to generate
corrected results. What's interesting is that the uncorrected results are
consistent with our previous findings and those from the
Right, strictly speaking, permutationis for orthogonal design matrices
only, which is not the case when you have a covariate. There was a
recent paper by Anderson Winkler Permutation Inference for the General
Linear Model in NI in which he goes over various permutation methods
for
Thanks! It seems that there was a white space. I ran the permutation
command but now I got another error:
ERROR: design matrix is not orthogonal, cannot be used with permutation.
If this something you really want to do, run with --perm-force
Wed May 27 16:27:50 EDT 2015
ERROR: cannot find any csd