JISCMail - FSL Archives

Hi Matthew,

I am running a PPI analyses (2 runs each for R and L hippocampal ROIs) and am experiencing these same errors for a few random participants. The errors are occurring inconsistently for one out of the two runs and one out of the two ROIs. Of note, all the design matrices include motion regressors as confound EVs generated using fsl_motion_outliers. I'm including the stats error output for one of the subjects below.

It was mentioned in the previous message that there may be a newer version of FILM that may fix this issue? Any insight would be greatly appreciated. Thank you so much!

Best,

Sahana

Stats output:

mkdir -p custom_timing_files ; /gpfs/milgram/apps/hpc.rhel7/software/FSL/6.0.0-centos7_64/bin/fslFixText /gpfs/milgram/project/gee_dylan/candlab/data/behavioral/shapes/task_design_prepostdot/A651/Testing1_APLUS_predot_8dv.txt custom_timing_files/ev1.txt

mkdir -p custom_timing_files ; /gpfs/milgram/apps/hpc.rhel7/software/FSL/6.0.0-centos7_64/bin/fslFixText /gpfs/milgram/project/gee_dylan/candlab/data/behavioral/shapes/task_design_prepostdot/A651/Testing1_APLUS_postdot_8dv.txt custom_timing_files/ev2.txt

mkdir -p custom_timing_files ; /gpfs/milgram/apps/hpc.rhel7/software/FSL/6.0.0-centos7_64/bin/fslFixText /gpfs/milgram/project/gee_dylan/candlab/data/behavioral/shapes/task_design_prepostdot/A651/Testing1_AMINUS_predot_8dv.txt custom_timing_files/ev3.txt

mkdir -p custom_timing_files ; /gpfs/milgram/apps/hpc.rhel7/software/FSL/6.0.0-centos7_64/bin/fslFixText /gpfs/milgram/project/gee_dylan/candlab/data/behavioral/shapes/task_design_prepostdot/A651/Testing1_AMINUS_postdot_8dv.txt custom_timing_files/ev4.txt

mkdir -p custom_timing_files ; /gpfs/milgram/apps/hpc.rhel7/software/FSL/6.0.0-centos7_64/bin/fslFixText /gpfs/milgram/project/gee_dylan/candlab/data/behavioral/shapes/task_design_prepostdot/A651/Testing1_BMINUS_predot_8dv.txt custom_timing_files/ev5.txt

mkdir -p custom_timing_files ; /gpfs/milgram/apps/hpc.rhel7/software/FSL/6.0.0-centos7_64/bin/fslFixText /gpfs/milgram/project/gee_dylan/candlab/data/behavioral/shapes/task_design_prepostdot/A651/Testing1_BMINUS_postdot_8dv.txt custom_timing_files/ev6.txt

mkdir -p custom_timing_files ; /gpfs/milgram/apps/hpc.rhel7/software/FSL/6.0.0-centos7_64/bin/fslFixText /gpfs/milgram/project/gee_dylan/candlab/data/behavioral/shapes/task_design_prepostdot/A651/Testing1_ABMINUS_predot_8dv.txt custom_timing_files/ev7.txt

mkdir -p custom_timing_files ; /gpfs/milgram/apps/hpc.rhel7/software/FSL/6.0.0-centos7_64/bin/fslFixText /gpfs/milgram/project/gee_dylan/candlab/data/behavioral/shapes/task_design_prepostdot/A651/Testing1_ABMINUS_postdot_8dv.txt custom_timing_files/ev8.txt

mkdir -p custom_timing_files ; /gpfs/milgram/apps/hpc.rhel7/software/FSL/6.0.0-centos7_64/bin/fslFixText /gpfs/milgram/project/gee_dylan/candlab/data/behavioral/shapes/task_design_prepostdot/A651/Testing1_ACMINUS_predot_8dv.txt custom_timing_files/ev9.txt

mkdir -p custom_timing_files ; /gpfs/milgram/apps/hpc.rhel7/software/FSL/6.0.0-centos7_64/bin/fslFixText /gpfs/milgram/project/gee_dylan/candlab/data/behavioral/shapes/task_design_prepostdot/A651/Testing1_ACMINUS_postdot_8dv.txt custom_timing_files/ev10.txt

mkdir -p custom_timing_files ; /gpfs/milgram/apps/hpc.rhel7/software/FSL/6.0.0-centos7_64/bin/fslFixText /gpfs/milgram/project/gee_dylan/candlab/data/behavioral/shapes/task_design_prepostdot/A651/Testing1_expectancy_8dv.txt custom_timing_files/ev11.txt

mkdir -p custom_timing_files ; /gpfs/milgram/apps/hpc.rhel7/software/FSL/6.0.0-centos7_64/bin/fslFixText /gpfs/milgram/project/gee_dylan/candlab/analyses/shapes/Adult_Control_n58/Level_1/A651_shapes2.feat/PPI/mean_timecourse_Hindy_R_AntHipp.txt custom_timing_files/ev12.txt

paste -d  ' '  /gpfs/milgram/project/gee_dylan/candlab/data/mri/hcp_pipeline_preproc/shapes/sub-A651/MNINonLinear/Results/ses-shapesV1_task-shapes2_bold/Confound_EVs_FDbxpltoutliers_MovReg_8dv_rawdata.txt > confoundevs.txt

/gpfs/milgram/apps/hpc.rhel7/software/FSL/6.0.0-centos7_64/bin/feat_model design confoundevs.txt

/gpfs/milgram/apps/hpc.rhel7/software/FSL/6.0.0-centos7_64/bin/film_gls --in=filtered_func_data --rn=stats --pd=design.mat --thr=11.628004 --sa --ms=5 --con=design.con  
Log directory is: stats
paradigm.getDesignMatrix().Nrows()=520
paradigm.getDesignMatrix().Ncols()=61
sizeTS=520
numTS=239825
Calculating residuals...
Completed
Estimating residual autocorrelation...
Calculating raw AutoCorrs... Completed
mode = 11065
sig = 1094
Spatially smoothing auto corr estimates
.....................
Completed
Tukey M = 22
Tukey estimates... Completed
Completed
Prewhitening and Computing PEs...
Percentage done:
1,2,3,4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19,20,21,22,23,24,25,26,27,
error: pinv(): svd failed

Uncaught exception!

FATAL ERROR ENCOUNTERED:
COMMAND:
/gpfs/milgram/apps/hpc.rhel7/software/FSL/6.0.0-centos7_64/bin/film_gls --in=filtered_func_data --rn=stats --pd=design.mat --thr=11.628004 --sa --ms=5 --con=design.con  
ERROR MESSAGE:
child process exited abnormally
END OF ERROR MESSAGE

On Thu, Apr 2, 2020 at 9:21 AM Sebastian Markett <[log in to unmask]> wrote:

Hi Matthew,

I think I am encountering a similar problem:

Running FEAT via the HCP pipelines the analysis aborts during Prewhitening and Computing PEs, throwing pinv(): svd failed as error message.

Strangely enough, this happens only for a fraction of my participants and selectively for some sessions, depending on the extend of spatial smoothing (2 mm FWHM error & 4 mm fine, for other runs the other way round). There doesn't seem to be any consistency.

You mentioned a fix for this and / or a newer version of FILM. You think I should give it a go?

Thanks a lot!
Sebastian

########################################################################

To unsubscribe from the FSL list, click the following link:
https://www.jiscmail.ac.uk/cgi-bin/webadmin?SUBED1=FSL&A=1