slow_surf_clustsim.py fails

AFNI Message Board
1

Dear AFNI experts,

I am running stats in AFNI on gyrification index images calculated with Freesurfer. I would like to perform correction for multiple comparisons on the results of the t-test and I am having some problems with slow_surf_clustsim.py to determine the minimum surface size to threshold my results.

The command I am running is
slow_surf_clustsim.py -save_script surf.sim.15_10iter -on_surface yes -uvar spec_file …/…/fsaverage_afni/SUMA/fsaverage_afni_lh.spec -uvar surf_vol …/…/fsaverage_afni/SUMA/fsaverage_afni_SurfVol+orig. -uvar blur 15.4 -uvar niter 10 -uvar results_dir ./clust_sim_15_10iter

The command correctly writes the script surf.sim.15_10iter, but when I run it I do not get the z.max.pvalue files necessary to determine the minimum surface size for the uncorrected p-value. It seems to me the somehow the command is not able to recognize the list of default p-values. These are the last lines of what I get in the terminal, even though not an error:
…
Writing FWHM progression history to smooth.noise.0.gii.1D.smrec …

#Final smoothing parameters via master:
#Niter Sigma OutputFWHM
285 0.5212 -1.000

165.60user 0.09system 2:45.85elapsed 99%CPU (0avgtext+0avgdata 214184maxresident)k
0inputs+17096outputs (0major+55824minor)pagefaults 0swaps
3dmaskave -mask all_1.gii -sigma smooth.noise.0.gii
awk {print $2}
++ 3dmaskave: AFNI version=AFNI_16.2.06 (Jul 25 2016) [64-bit]
+++ 163842 voxels survive the mask
3dcalc -a smooth.noise.0.gii -b t.stdev.1D -expr a/b -prefix smooth.white.0.gii
++ 3dcalc: AFNI version=AFNI_16.2.06 (Jul 25 2016) [64-bit]
++ Authored by: A cast of thousands
++ Output dataset ./smooth.white.0.gii
@ iminus1 = 10 - 1
foreach index ( count -digits 1 0 $iminus1 )
count -digits 1 0 9
foreach pind ( count -digits 1 1 $#pthr_list )
count -digits 1 1 4
set pthr =
set zither =

/…/fsaverage_afni/SUMA/fsaverage_afni_lh.spec -surf_A smoothwm -input smooth.white.0.gii[0] 0 -sort_area -rmm -1 -athresh
need argument after -athresh
0.00user 0.00system 0:00.01elapsed 26%CPU (0avgtext+0avgdata 4228maxresident)k
0inputs+0outputs (0major+1174minor)pagefaults 0swaps

Any idea what is happening? This is the list of the files written in the folder:

-rw-r–r–. 1 30 Aug 11 13:50 dummy.TRs.10.1D
-rw-r–r–. 1 2.2M Aug 11 13:50 nodes.1D
-rw-r–r–. 1 855K Aug 11 13:50 empty.gii
-rw-r–r–. 1 855K Aug 11 13:50 all_1.gii
-rw-r–r–. 1 8.4M Aug 11 13:50 surf.noise.0.gii
-rw-r–r–. 1 7.6K Aug 11 13:53 smooth.noise.0.gii.1D.smrec
-rw-r–r–. 1 8.4M Aug 11 13:53 smooth.noise.0.gii
-rw-r–r–. 1 109 Aug 11 13:53 params.surf.smooth.0.1D
-rw-r–r–. 1 100 Aug 11 13:53 t.stdev.1D
-rw-r–r–. 1 8.4M Aug 11 13:53 smooth.white.0.gii
-rw-r–r–. 1 0 Aug 11 13:53 clust.out.0.0.

thank you very much for your help.

Best,
Giovanni

2

Hi Giovanni,

Your command is working for me. Please send me email with 2 files
(it is a bit much to post here): the actual surf.sim.15_10iter script, along
with the output of:

tcsh -x surf.sim.15_10iter |& tee out.surf.sim.15_10iter.txt

So mail me surf.sim.15_10iter and out.surf.sim.15_10iter.txt .
Clicking on my name should show the email address, I believe.

Thanks,

  • rick
3

Thanks for your help Rick. I have sent you an email with the files you requested.

Best,
Giovanni

4

Thanks Giovanni.

Your script it identical with one of my own, which is
working fine (but on my system and with my data).
But I see that your AFNI version might be 3 years
old. Is it?

Can you regenerate the output text? It just shows
that the result directory already exists. I guess you
can delete the old directory and re-run it.

Thanks,

  • rick
5

Hi Rick,
I tried to run the command with an updated version of AFNI and everything works fine.

Thank you very much for your help.

Best,
Giovanni

6

That is good news, thanks for the update!

  • rick