help > Problem with running gPPI
Showing 1-11 of 11 posts
Display:
Results per page:
Feb 2, 2020  11:02 PM | Gal N-Cohen - Ben-Gurion University of the Negev
Problem with running gPPI
Hi all, 
I'm trying to run gPPI and getting an unspecific error: 
"ERROR 1: PPPI inputs were specified correctly, but failed error checking for dependencies. This is likely a bug in the program.
One or more inputs are not correct."

Unfortunately, similar older forum posts weren't helpful here (problem with directory, etc.).

Here is the P. file as specified : 

subject: 's02'
VOI: 'C:\PhD DATA\gPPI\ROI\ffa_L.nii'
Region: 'ffa_L'
SPMver: 12
directory: 'C:\PhD DATA\gPPI\res\s02\Model4b_folder'
Estimate: 1
contrast: 0
extract: 'mean'
Tasks: {'a' 'b' 'c' 'd' 'e' 'f' 'g' 'h' 'blank' 'error'}
Weights: []
maskdir: []
equalroi: 0
FLmask: 1
VOI2: {}
analysis: 'psy'
method: 'cond'
CompContrasts: 1
Weighted: []
Contrasts: [1×2 struct]
Contrasts(1) = 
left: {'a' 'b'}
right: {'c' 'd'}
STAT: 'T'
Weighted: 0
MinEvents: 5
name: 'first_con'
Contrasts(2) = 
left: {'e' 'f'}
right: {'g' 'h'}
STAT: 'T'
Weighted: 0
MinEvents: 5
name: '
second_con'
TasksBackup: {'a' 'b' 'c' 'd' 'e' 'f' 'g' 'h' 'blank' 'error'}
zipfiles: 0
GroupDir: 'C:\PhD DATA\gPPI\res\ffa_L'
concatR: 1


Thanks in advance, 
Gal
Feb 3, 2020  05:02 AM | Donald McLaren
RE: Problem with running gPPI
Could you attach the log file?

It should be in "C:\PhD DATA\gPPI\res\s02\Model4b_folder" and end in ".log"

If there is no log file there, what is the contents of that directory?

Thanks.
Feb 3, 2020  07:02 AM | Gal N-Cohen - Ben-Gurion University of the Negev
RE: Problem with running gPPI
Thanks for your quick reply. 

This directory contains the SPM.mat file, the new generated P file (per region, per subject) and the attached log file.
On my first attempt, the log file was empty. On second attempt I've received the attached log file.  

Thanks,
Gal
Feb 3, 2020  08:02 AM | Donald McLaren
RE: Problem with running gPPI
Hi Gal,

Can you try changing P.Tasks from {'a' 'b' 'c' 'd' 'e' 'f' 'g' 'h' 'blank' 'error'} to {'0' 'a' 'b' 'c' 'd' 'e' 'f' 'g' 'h' 'blank' 'error'} and let me know if that solves the issue.

If that works, then you can decide if all tasks need to be present in all runs or not. If they need to be present, then change the 0 to 1.

If that doesn't work, please send me the SPM.mat file and I will try running through the program on my machine to replicate the issue.

Best,
Donald
Feb 3, 2020  08:02 AM | Gal N-Cohen - Ben-Gurion University of the Negev
RE: Problem with running gPPI
Thanks Donald, 
I tried changing the first cell to 0  as suggested and got the same error. Then, I changed the first cell from 0 to 1 and got other error ("Program will exit. Tasks are not defined in an array.").

I am attaching here the SPM.mat file

Thanks,
Gal
Attachment: spm.mat
Feb 3, 2020  12:02 PM | Donald McLaren
RE: Problem with running gPPI
Hi Gal,

You need to give your tasks character names. At present, they are numeric, which breaks the code. It also causes the SPM.xX.name to have repeating names as the numbers don't give the columns the correct names.

If you change the task names in the original model, then the code should work. Additionally, you need to make sure that P.Tasks array starts with '0' or '1' and that the task names entered here match those in the SPM model.

If all the task names match and you still have an issue, please upload the revised SPM.mat and I will take a deeper look.

Best,
Donald
Originally posted by Gal N-Cohen:
Thanks Donald, 
I tried changing the first cell to 0  as suggested and got the same error. Then, I changed the first cell from 0 to 1 and got other error ("Program will exit. Tasks are not defined in an array.").

I am attaching here the SPM.mat file

Thanks,
Gal
Feb 3, 2020  11:02 PM | Gal N-Cohen - Ben-Gurion University of the Negev
RE: Problem with running gPPI
Hi Donald, 
I tried it and generated a new spm.mat file (by running a new 1st level analysis; attached). This time I named the tasks with letters rather than numbers. 
Unfortunately, It still doesn't work.

I really appreciate your help,
Thanks,
Gal
Feb 4, 2020  01:02 AM | Gal N-Cohen - Ben-Gurion University of the Negev
RE: Problem with running gPPI
Originally posted by Gal N-Cohen:
Hi Donald, 
I tried it and generated a new spm.mat file (by running a new 1st level analysis; attached). This time I named the tasks with letters rather than numbers. 
Unfortunately, It still doesn't work.

I really appreciate your help,
Thanks,
Gal
Attachment: SPM.mat
Feb 9, 2020  03:02 AM | Gal N-Cohen - Ben-Gurion University of the Negev
RE: Problem with running gPPI
Hi Donald,
Is there a chance that the gPPI code is not working due to my SPM version? 
I changed the PPPI.m line to SPM12 rather than 8, but might it be that there are some SPM8 functions that not been implemented in SPM12 and required for the gPPI analysis? 
I can't find why it keeps failing. 

Thanks again,
Gal
Feb 19, 2020  05:02 PM | Donald McLaren
RE: Problem with running gPPI
Hi Gal,

Sorry for the delay in the response.

Please change 

P.Tasks={'a' 'b' 'c' 'd' 'e' 'f' 'g' 'h' 'blank' 'error'};

to

P.Tasks={'1' 'a' 'b' 'c' 'd' 'e' 'f' 'g' 'h' 'blank' 'error'}; or P.Tasks={'0' 'a' 'b' 'c' 'd' 'e' 'f' 'g' 'h' 'blank' 'error'}; 

depending on if every task needs to be in every run.

Once you have the new P-file set, try running P=PPPI_checkstruct(P). If you get no error messages, then you should be okay.

If it fails after this step, I'll have to think about how to test it without your exact dataset.

Does the test dataset work? This would determine if it's dataset specific or a more general issue?
Feb 27, 2020  08:02 AM | Gal N-Cohen - Ben-Gurion University of the Negev
RE: Problem with running gPPI
Hi Donald, thank you for

I tried using "0" and "1" at the beginning (and I tried it with P.Tasks = {}) and it works. However, I have a new problem now and it is the same error on the sample dataset and on my data. The gPPI is running but there is a problem with the mask\ROI: 

PPPI Version: 13.1.4-17-2014
Parameters used:
Processing subject: sampleData_Study
VOI file : C:\PhD DATA\gPPI\NIV\sample dataset\running_gPPI_generic\rmedPrecCing.nii
Output file will be: sampleData_Study_rmedPrec_session#_cond_PPI_regressors.txt
Contrast : No adjustment
Analysis : Psychophysiological Interactions
Extraction : eigenvariate
Tasks : _0_etoh_on_onsets_run1_etoh_off_onsets_run1_etoh_urge_onsets_run1_attn_on_onsets_run1_attn_off_onsets_run1_attn_urge_onsets_run1_etoh_on_onsets_run2_etoh_off_onsets_run2_etoh_urge_onsets_run2_attn_on_onsets_run2_attn_off_onsets_run2_attn_urge_onsets_run2
Method : Condition Specific


Undefined function or variable 'set_mask'.

Error in timeseries_extract (line 257)
Error in PPPI (line 461)

I've tried to search by myself, step by step, at the timeseries_extract code, but I couldn't figure how to solve the problem. 
The "checkstruct" is OK, I haven't got any error messages. 

 Thank you so much! 
Gal