Dear Buyean,
Many thanks for pointing this out - it is actually a serious bug.
I've uploaded a patch to concat_bvars which fixes it:
http://www.fmrib.ox.ac.uk/fsldownloads/patches/
but this does have consequences for any previous analysis run with
concat_bvars, as it produced an alphabetically ordered list of files
internally. If you did have an alphabetically ordered list (e.g.
single wildmask)
then there is no problem. Also, if you had equal group sizes with
consistent
naming then all stats are likely to be fine, but vectors and surfaces
may be
reversed. Otherwise, things could be quite different.
I would recommend re-running concat_bvars with the new version
to create a new file, and if it is different from the old one (check
with:
diff file1 file2) then you should re-run any vertex analysis that used
this.
Many apologies for this bug.
I hope that it does not cause major problems for anyone.
It has existed since FSL4.1.4 (22 May 2009).
Sorry,
Mark
On 2 Sep 2009, at 03:33, Buyean Lee wrote:
> Hi,
>
> I expected to see the direction of the arrows in vertex analysis
> results to be changed opposite when I swapped the order of *.bvars
> files during concatenation (e.g., from (1) con_to_dis to (2)
> dis_to_con, as below) and kept the design matrix file the same.
>
> (1) concat_bvars con_to_dis_L_Hipp.bvars con*_L_Hipp_first.bvars
> dis*_L_Hipp_first.bvars
> (2) concat_bvars dis_to_con_L_Hipp.bvars dis*_L_Hipp_first.bvars
> con*_L_Hipp_first.bvars
>
> However, I directions of the arrows of the two vertex analyses are
> the same (pointing inward).
> In contrast, when I swapped the 0 and 1 in the design matrix file
> with the same concatenated bvars file, the directions of the arrows
> were changed.
> I wonder if this is normal or I did something wrong.
>
> Thank you,
>
> Buyean
|