Hi Florencia,It turns out that the mask includes voxels with constant values, that is, with zero variance. Then these voxels overlap with the intercept that is already present in the model.For data entered with -i, these voxels are automatically dropped, but not with data entered with -evperdat. The reason is to allow the user to, in some cases, have an intercept there. If these voxels had zero mean, then they would be removed.Since this is your sole regressor of interest, and there are no nuisance variables other than the intercept, the solution is simple: drop the design (i.e., remove "-d Design.mat"), and drop the "2" from the -evperdat, so that it falls back to "1" which is the default EV position, and finally, add "-demean". It should work then.Hope this helps!All the best,AndersonSure! I'll send them right away off-list.Thanks for the help!Hi Florencia,Could you send me your input files (desing, mask, contrasts, niftis, etc)? You can place them in a file sharing service and send the link off-list.Thanks!All the best,AndersonHi Anderson, I have updated to alpha116 and tried rerunning the commandspalm -i DTI_4D_S1-S2.nii -i DTI_4D_S3-S2.nii -evperdat 4d_BOLD.nii -npc -npcmod -corrcon -corrmod -n 500 -T -Tnpc -logp -accel tail -concordant -nouncorrected -o $output -m $maskandpalm -i DTI_4D_S1-S2.nii -i DTI_4D_S3-S2.nii -evperdat 4d_BOLD.nii 2 -d Design_voxelwiseEV.mat -t Design_voxelwiseEV.con -npc -npcmod -corrcon -corrmod -n 500 -T -Tnpc -logp -accel tail -nouncorrected -o $output -m $maskUnfortunately, I am still getting the same warnings.Hi Florencia,PALM alpha116 has just been released. Maybe give it a try? Please let me know if the problem persists.All the best,AndersonHi Florencia,Please see below:Dear Anderson and FSL users,I am trying to run an analysis to relate changes in DTI with BOLD (BOLD and DTI images arein the same space and have the same spatial resolution). Thus, I have used the BOLD image as voxelwise EV in PALM. I have tried two options and both present problems:1) Use BOLD as voxelwise EV for DWI changes with FWE correction across contrasts and modalities and NPC across modalities
palm -i DTI_4D_S1-S2.nii -i DTI_4D_S3-S2.nii -evperdat 4d_BOLD.nii -npc -npcmod -corrcon -corrmod -n 500 -T -Tnpc -logp -accel tail -concordant -nouncorrected -o $output -m $maskIn this case, I am getting the following warning:Rank deficient = 0I have tried adding -pmethodr Ridgway as partition method used in the regression to solve the rank deficiency, as I had seen suggested in a thread in this forum, but it didn't work for me.For the voxelwise EV (-evperdat) this is a hideous bug that was fixed yesterday. However, the change hasn't been committed yet. I'll try to post to Github later today.2) Use BOLD as voxelwise EV for DWI changes with FWE correction across contrasts and modalities and design. The design I used is all ones in EV1 for the mean and all ones in EV2 for the voxelwise EV. The contrasts I used are [0 1] and [0 -1] to test for positive and negative associations.palm -i DTI_4D_S1-S2.nii -i DTI_4D_S3-S2.nii -evperdat 4d_BOLD.nii 2 -d Design_voxelwiseEV.mat -t Design_voxelwiseEV.con -npc -npcmod -corrcon -corrmod -n 500 -T -Tnpc -logp -accel tail -nouncorrected -o $output -m $maskIn this case, I am getting a warning saying Matrix is singular to working precision.What could be the problem here?I believe this is the same problem as above. However, it'd help of you could paste the full error message (with line numbers) so this can be checked more accurately...Thanks, and apologies for the long delay...All the best,AndersonThanks for any advice you can give me to solve this.Regards,Florencia.--Ing. Florencia Jacobacci.PhD student / Becaria doctoralInstituto de Fisiología y Biofísica (IFIBIO) - Bernardo HoussayLaboratorio de Fisiología de la Acción, Facultad de MedicinaUniversidad de Buenos AiresParaguay 2155, C.A.B.A.(C1121ABG), Argentina☎ (+5411) 5285-3304
To unsubscribe from the FSL list, click the following link:
https://www.jiscmail.ac.uk/cgi-bin/webadmin?SUBED1=FSL&A=1
To unsubscribe from the FSL list, click the following link:
https://www.jiscmail.ac.uk/cgi-bin/webadmin?SUBED1=FSL&A=1--Ing. Florencia Jacobacci.PhD student / Becaria doctoralInstituto de Fisiología y Biofísica (IFIBIO) - Bernardo HoussayLaboratorio de Fisiología de la Acción, Facultad de MedicinaUniversidad de Buenos AiresParaguay 2155, C.A.B.A.(C1121ABG), Argentina☎ (+5411) 5285-3304
To unsubscribe from the FSL list, click the following link:
https://www.jiscmail.ac.uk/cgi-bin/webadmin?SUBED1=FSL&A=1
To unsubscribe from the FSL list, click the following link:
https://www.jiscmail.ac.uk/cgi-bin/webadmin?SUBED1=FSL&A=1
--Ing. Florencia Jacobacci.PhD student / Becaria doctoralInstituto de Fisiología y Biofísica (IFIBIO) - Bernardo HoussayLaboratorio de Fisiología de la Acción, Facultad de MedicinaUniversidad de Buenos AiresParaguay 2155, C.A.B.A.(C1121ABG), Argentina☎ (+5411) 5285-3304
To unsubscribe from the FSL list, click the following link:
https://www.jiscmail.ac.uk/cgi-bin/webadmin?SUBED1=FSL&A=1
To unsubscribe from the FSL list, click the following link:
https://www.jiscmail.ac.uk/cgi-bin/webadmin?SUBED1=FSL&A=1
To unsubscribe from the FSL list, click the following link:
https://www.jiscmail.ac.uk/cgi-bin/webadmin?SUBED1=FSL&A=1