help
47 Subscribers 
help > Problem with spm_estimate PPI
Showing 1-4 of 4 posts
Mar 27, 2018 09:03 AM | Konstantinos Bromis
Problem with spm_estimate PPI
Hi,
When I run gPPI analysis I get the error below:
Error using load
'stats_right_hippocampus_session1_cond_PPI_regressors.mat' is not found in the current folder or on the MATLAB path.
However, the above error comes when I run additionally the sentence below:
SPM = spm_estimate_PPI(P.subject,fullfile(directories.rdir,directories.sdir,'SPM.mat'),P.Region,P.method,P.analysis,P.CompContrasts,0,[],0,0)
When I turn this to a comment and run the script again I get an error of estimation failure Warning: Missing conditions!!! Invalid Contrast.
However, at this moment the 'stats_right_hippocampus_session1_cond_PPI_regressors.mat' has been created. So, if I run the script as I did at first (by using the SPM=spm_estimate_PPI(P.subject,fullfile(directories.rdir,directories.sdir,'SPM.mat'),P.Region,P.method,P.analysis,P.CompContrasts,0,[],0,0) everything goes well.
Hovever, I have to do this manually everytime for each subject and its time consuming.
Is there any way it could be fixed?
Thank you.
Konstantinos
When I run gPPI analysis I get the error below:
Error using load
'stats_right_hippocampus_session1_cond_PPI_regressors.mat' is not found in the current folder or on the MATLAB path.
However, the above error comes when I run additionally the sentence below:
SPM = spm_estimate_PPI(P.subject,fullfile(directories.rdir,directories.sdir,'SPM.mat'),P.Region,P.method,P.analysis,P.CompContrasts,0,[],0,0)
When I turn this to a comment and run the script again I get an error of estimation failure Warning: Missing conditions!!! Invalid Contrast.
However, at this moment the 'stats_right_hippocampus_session1_cond_PPI_regressors.mat' has been created. So, if I run the script as I did at first (by using the SPM=spm_estimate_PPI(P.subject,fullfile(directories.rdir,directories.sdir,'SPM.mat'),P.Region,P.method,P.analysis,P.CompContrasts,0,[],0,0) everything goes well.
Hovever, I have to do this manually everytime for each subject and its time consuming.
Is there any way it could be fixed?
Thank you.
Konstantinos
Mar 28, 2018 12:03 AM | Donald McLaren
RE: Problem with spm_estimate PPI
Hi,
Could you kindly attach your script, so I can see where the these variables are being generated? My copy of PPPI.m, which is the core toolbox function does not use fullfile anywhere.
Best,
Donald
Could you kindly attach your script, so I can see where the these variables are being generated? My copy of PPPI.m, which is the core toolbox function does not use fullfile anywhere.
Best,
Donald
Mar 28, 2018 10:03 AM | Konstantinos Bromis
RE: Problem with spm_estimate PPI
Hi,
I have attached the script.
Thank you.
Konstantinos
I have attached the script.
Thank you.
Konstantinos
Mar 28, 2018 02:03 PM | Donald McLaren
RE: Problem with spm_estimate PPI
Hi Konstantinos,
I assume that you've already estimated your first level task model. If this is the case, running the command without line 59 SPM=spm_estimate_ppi... should solve the issue of it running.
PPPI(P) will estimate the model and contrasts for you.
However, I don't believe that the PPPI toolbox can handle .gz VOI files. Could you change this file to simply be a .nii file and run the script again?
If you get the estimation failure again, please copy and past the full matlab window output along with stats_right_hippocampus_session1_cond_PPI_regressors.mat file.
I suspect the issue maybe that the stats_right_hippocampus_session1_cond_PPI_regressors.mat file is empty and thus there are no PPI columns in the model. So the model estimates, but the contrasts can't find the PPI columns.
Best,
Donald
I assume that you've already estimated your first level task model. If this is the case, running the command without line 59 SPM=spm_estimate_ppi... should solve the issue of it running.
PPPI(P) will estimate the model and contrasts for you.
However, I don't believe that the PPPI toolbox can handle .gz VOI files. Could you change this file to simply be a .nii file and run the script again?
If you get the estimation failure again, please copy and past the full matlab window output along with stats_right_hippocampus_session1_cond_PPI_regressors.mat file.
I suspect the issue maybe that the stats_right_hippocampus_session1_cond_PPI_regressors.mat file is empty and thus there are no PPI columns in the model. So the model estimates, but the contrasts can't find the PPI columns.
Best,
Donald