Dear all,

I am experiencing an odd combination of errors when running gPPI (psychophysiological interaction) in the context of a fairly complex study design. I have 36 tasks across 4 runs (2 sessions each containing 2 runs). Each run contains 9 tasks. When specifying my P.Task, I therefore encode all tasks with '0'. I am using a localising contrast and a sphere ROI created in FSLmaths based on the normalising MNI template. (This ROI is in .nii format, which I understand to be one of the acceptable types.) I then create a contrast using two tasks from one run and two from another.

(I plan on re-running using an equivalent localising contrast and the same ROI. Then contrasting a different pair of two tasks on the left-hand and two tasks on the right-hand. This will effectively recapitulate the gPPI analysis at the first-level for two levels of a factorial variable, using specialised localisers. Comparison between these levels will be possible at the second-level.)

One complicating factor is that separate images for each session exist, however these are pointed to by the same first-level GLM SPM.

When I run the attention sample PPI everything works out fine. I have resolved the known dlmwrite and spm version issues (running in SPM12).

I think the errors are somewhat conflicting. Error 1 suggests that the number of tasks exceeds that in the GLM. However we have 36 task regressors and 4 nuisance (nonconvolved regressors) in the GLM, checked with spm.xX.name. I name all 36 of the task regressors in P.Task. Error 2 suggests that the parameters.mat is acceptable, but that there are downstream errors.

I am concerned that my design may be so complex, in terms of the number of task regressors that it is breaking the software. I chose to use your toolbox as I felt that (i) it allows contrasts performed to be restricted to just those specified, rather than all possible and (ii) standard sPPI is undesirable due to only using 4/36 task regressors and therefore 4/36th of the experimental space at the first-level. Can anyone advise?

ERROR 1: Program will now exit. You specified too many tasks. At least one specified task does not exist.
ERROR 2: PPPI inputs were specified correctly, but failed error checking for dependencies. See errorval{1}.
One or more inputs are not correct.
ERROR 1:Program will now exit. You specified too many tasks. At least one specified task does not exist.
ERROR 2:PPPI inputs were specified correctly, but failed error checking for dependencies. See errorval{1}.
errorvals saved to errorvals.mat

Many thanks,

Mark

Dear All,

I think that I have localised this error to a failure to load the first-level SPM.mat GLM data correctly. This is apparent since the Tasks variable isn't found after running the PPPI(P) command, though the parameter variable P.Tasks is found. This is consistent with the errors generated by PPPIinputsvalid and PPPIcheckstruct.

None of the SPM variables are not loaded into the workspace suggesting that there is a general or partial failure to load the SPM.mat, rather than a specific failure in getting the Tasks variable to load.

Many thanks,

Mark


PPPIinputsvalid.m
elseif ~isempty(P.Tasks) && str2double(P.Tasks{1})==0 && (length(P.Tasks)-1)>numel(Tasks)
error='Program will now exit. You specified too many tasks. At least one specified task does not exist.';

PPPI_checkstruct.m
catch
error{end+1}='PPPI inputs were specified correctly, but failed error checking for dependencies. See errorval{1}.';
end

Dear All,

I think that I have localised this error to a failure to load the first-level SPM.mat GLM data correctly. This is apparent since the Tasks variable isn't found after running the PPPI(P) command, though the parameter variable P.Tasks is found. This is consistent with the errors generated by PPPIinputsvalid and PPPIcheckstruct.

None of the SPM variables are not loaded into the workspace suggesting that there is a general or partial failure to load the SPM.mat, rather than a specific failure in getting the Tasks variable to load.

Many thanks,

Mark


PPPIinputsvalid.m
elseif ~isempty(P.Tasks) && str2double(P.Tasks{1})==0 && (length(P.Tasks)-1)>numel(Tasks)
error='Program will now exit. You specified too many tasks. At least one specified task does not exist.';

PPPI_checkstruct.m
catch
error{end+1}='PPPI inputs were specified correctly, but failed error checking for dependencies. See errorval{1}.';
end