quote:
Greetings All,
We've encountered a very odd problem with behavioral PLS on a PET dataset in Matlab 7 (and 6.5). When we run the PLS analysis permutations always go through but the process stops at a random bootstrap (in this case 7, we set permutations to 500 and bootstrap to 100). The error follows:
>> plsgui
??? Attempted to access reorderp(:,7); index out of bounds because size(reorderp)=[14,1].
Error in ==> pet_analysis_boot at 325
reorder(:,p) = reorderp(:,p);
Error in ==> pet_analysis at 243
[brainlv2,s2,behavlv2,brainscores2,behavscores2,lvcorrs2, ...
Error in ==> pet_analysis_ui>ExecutePLS at 1881
[resultFile, elapsed_time] = pet_analysis(PLSoptions.isbehav, ...
Error in ==> pet_analysis_ui at 114
ExecutePLS;
??? Error using ==> waitfor
Error while evaluating uicontrol Callback.
This is our behavioral data matrix - 14 subjects / 1 condition (from the CSV):
120,2,2,4,5,0,0,5,0,0,3,4,2
120,1,1,0,5,2,0,3,2,0,0,3,0
120,2,1,3,13,1,0,11,3,1,0,3,0
99,14,1,5,19,1,0,12,0,0,0,5,0
120,5,0,4,3,1,0,6,1,0,0,5,1
57,14,0,5,6,0,0,5,0,0,6,5,0
52,14,1,5,5,0,0,5,0,0,10,5,0
120,7,3,6,17,2,0,14,1,0,6,5,0
120,2,0,3,6,0,0,12,0,0,4,3,2
120,2,2,3,3,2,2,12,1,0,2,4,0
73,14,0,4,11,2,0,11,1,1,14,9,8
120,3,3,2,1,0,1,10,0,0,2,6,5
73,14,1,3,7,0,0,8,0,0,3,7,5
120,3,0,2,9,2,0,8,0,0,0,4,0
Even on the same computer, for the exact same datamat it will stop on a different (though close) bootstrap. In addition, we had run an analysis with some old images under the same setup and that worked just fine. This set is much smaller than what we are typically able to run, so we do not think it is a resource issue. Any explanations or help is appreciated.
Many Thanks,
Rayhan and the Jagust Lab
Rayhan..
One possibility -
If I understand correctly, you are looking at all 13 behavioural variables in this analysis?
If so, there is the possibility that the bootstrap "reorders" could not be calculated properly because of the low variance in some of the columns. We check for zero variance and resample.
Can you try it again, using just columns 1,4,5,8..
thanks
Nancy