afni_proc.py group mask and -tlrc_NL_warped_dsets

AFNI Message Board
1

Hi,

I’m trying to rerun some RSFC analysis and in doing so also trying to avoid rerunning nonlinear warping given how long it takes.

I’m running the following command from linux_openmp_64: Aug 19 2016 (Version AFNI_16.2.11)


 afni_proc.py -subj_id 106_A -script                                                                       \
      106_A.afniRsfcPreprocess.noanaticor.NL.csh -out_dir                                                   \
      afniRsfcPreprocessed.noanaticor.NL -blocks despike tshift align tlrc                                  \
      volreg mask blur regress -copy_anat                                                                   \
      /data/sanDiego/machLearnT1Analysis/data/106_A/106_A.anat.zp+orig.HEAD                                 \
      -dsets                                                                                                \
      /data/sanDiego/machLearnT1Analysis/data/106_A/106_A.resting.zp+orig.HEAD                              \
      -tlrc_base MNI_caez_N27+tlrc -volreg_align_to first -volreg_tlrc_warp                                 \
      -align_opts_aea -cost lpc -tlrc_NL_warp -tlrc_NL_warped_dsets                                         \
      /data/sanDiego/machLearnT1Analysis/data/106_A/afniRsfcPreprocessed.NL/106_A.anat.zp_al_keep+tlrc.HEAD \
      /data/sanDiego/machLearnT1Analysis/data/106_A/afniRsfcPreprocessed.NL/anat.un.aff.Xat.1D              \
      /data/sanDiego/machLearnT1Analysis/data/106_A/afniRsfcPreprocessed.NL/anat.un.aff.qw_WARP.nii         \
      -blur_size 7 -blur_to_fwhm -blur_opts_B2FW '-ACF -rate 0.2 -temper'                                   \
      -mask_apply group -mask_segment_anat yes -regress_censor_first_trs 3                                  \
      -mask_segment_erode yes -regress_ROI WMe -regress_bandpass 0.01 0.1                                   \
      -regress_apply_mot_types demean -regress_censor_motion 0.25                                           \
      -regress_run_clustsim no -regress_est_blur_epits -regress_est_blur_errts

but am getting the following error


-- tcat: reps is now 256
++ updating polort to 4, from run len 512.0 s
-- importing NL-warp datasets
-- volreg: using base dset vr_base+orig
++ volreg: applying volreg/tlrc xforms to isotropic 3 mm tlrc voxels
-- applying anat warps to 1 dataset(s): 106_A.anat.zp
** ERROR: cannot apply group mask
** script creation failure for block 'mask'
-- will apply motion types: demean
** regress_ROI: missing ROI dset for 'WMe'
** script creation failure for block 'regress'

Is it not possible to supply previously computed nonlinear warps and apply a group mask?

Thanks,

Colm.

2

Hi Colm,

Okay, I will fix that and fire off a build.
Binaries will be ready in about 2 hours.

Thanks!

  • rick
3

Much appreciated, thanks!

4

Thanks for the fix Rick. This now seems to work to a point. afni_proc will now generate a shell script but running it unmasks another error (all 3 of the pre-warped files exist):


 afni_proc.py -subj_id 107_A -script 107_A.afniRsfcPreprocess.noanaticor.NL.csh -out_dir afniRsfcPreprocessed.noanaticor.NL -blocks despike tshift align tlrc volreg mask blur regress -copy_anat /data/sanDiego/machLearnT1Analysis/data/107_A/107_A.anat+orig.HEAD -dsets /data/sanDiego/machLearnT1Analysis/data/107_A/107_A.resting+orig.HEAD -tlrc_base MNI_caez_N27+tlrc -volreg_align_to first -volreg_tlrc_warp -tlrc_NL_warp -tlrc_NL_warped_dsets /data/sanDiego/machLearnT1Analysis/data/107_A/afniRsfcPreprocessed.NL/107_A.anat_al_keep+tlrc.HEAD /data/sanDiego/machLearnT1Analysis/data/107_A/afniRsfcPreprocessed.NL/anat.un.aff.Xat.1D /data/sanDiego/machLearnT1Analysis/data/107_A/afniRsfcPreprocessed.NL/anat.un.aff.qw_WARP.nii -blur_size 7 -blur_to_fwhm -blur_opts_B2FW '-ACF -rate 0.2 -temper' -mask_apply group -mask_segment_anat yes -regress_censor_first_trs 3 -mask_segment_erode yes -regress_ROI WMe -regress_bandpass 0.01 0.1 -regress_apply_mot_types demean -regress_censor_motion 0.25 -regress_run_clustsim no -regress_est_blur_epits -regress_est_blur_errts
-- tcat: reps is now 256
++ updating polort to 4, from run len 512.0 s
-- importing NL-warp datasets
-- volreg: using base dset vr_base+orig
++ volreg: applying volreg/tlrc xforms to isotropic 3 mm tlrc voxels
-- applying anat warps to 1 dataset(s): 107_A.anat
-- masking: group anat = 'MNI_caez_N27+tlrc', exists = 1
-- will apply motion types: demean
-- have 1 ROIs to regress: WMe

-------------------------------------
** warning have only 1 run to analyze
-------------------------------------

--> script is file: 107_A.afniRsfcPreprocess.noanaticor.NL.csh
    consider the script execution command: 

      tcsh -xef 107_A.afniRsfcPreprocess.noanaticor.NL.csh |& tee output.107_A.afniRsfcPreprocess.noanaticor.NL.csh

+ [[ -f 107_A.afniRsfcPreprocess.noanaticor.NL.csh ]]
+ tcsh -xef 107_A.afniRsfcPreprocess.noanaticor.NL.csh
echo auto-generated by afni_proc.py, Thu Aug 25 17:59:46 2016
auto-generated by afni_proc.py, Thu Aug 25 17:59:46 2016
echo (version 5.02, August 25, 2016)
(version 5.02, August 25, 2016)
echo execution started: `date`
date
execution started: Thu Aug 25 17:59:46 PDT 2016
afni -ver
Precompiled binary linux_openmp_64: Aug 25 2016 (Version AFNI_16.2.13)
afni_history -check_date 1 Dec 2015
-- is current: afni_history as new as: 01 Dec 2015
               most recent entry is:   25 Aug 2016
if ( 0 ) then
if ( 0 > 0 ) then
set subj = 107_A
endif
set output_dir = afniRsfcPreprocessed.noanaticor.NL
if ( -d afniRsfcPreprocessed.noanaticor.NL ) then
set runs = ( `count -digits 2 1 1` )
count -digits 2 1 1
mkdir afniRsfcPreprocessed.noanaticor.NL
mkdir afniRsfcPreprocessed.noanaticor.NL/stimuli
3dcopy 107_A.anat+orig afniRsfcPreprocessed.noanaticor.NL/107_A.anat
++ 3dcopy: AFNI version=AFNI_16.2.13 (Aug 25 2016) [64-bit]
3dcopy afniRsfcPreprocessed.NL/107_A.anat_al_keep+tlrc afniRsfcPreprocessed.noanaticor.NL/107_A.anat_al_keep
++ 3dcopy: AFNI version=AFNI_16.2.13 (Aug 25 2016) [64-bit]
3dcopy afniRsfcPreprocessed.NL/anat.un.aff.Xat.1D afniRsfcPreprocessed.noanaticor.NL/anat.un.aff.Xat.1D
++ 3dcopy: AFNI version=AFNI_16.2.13 (Aug 25 2016) [64-bit]
3dcopy afniRsfcPreprocessed.NL/anat.un.aff.qw_WARP.nii afniRsfcPreprocessed.noanaticor.NL/anat.un.aff.qw_WARP.nii
++ 3dcopy: AFNI version=AFNI_16.2.13 (Aug 25 2016) [64-bit]
** ERROR (nifti_image_write_hdr_img2): cannot open output file 'afniRsfcPreprocessed.noanaticor.NL/afniRsfcPreprocessed.noanaticor.NL/anat.un.aff.qw_WARP.nii.gz'
5

Hi Colm,

I would have to guess that you have AFNI_AUTOGZIP set. That
variable causes a lot of difficulty with .nii NIFTI datasets, because
they get changed to .nii.gz. Can you verify whether that variable
is set?

Thanks,

  • rick
6

Yeah, I always have it set to reduce disk usage. Is the only way around it to unset it?

7

The alternative would probably be to convert the
NIFTI dataset to AFNI, and then it would not matter.

  • rick