Back to PLS Helperror "Too much data"
Posted on 08/10/12 17:38:46
Number of posts: 6
Hi,
I plan to do a behavioral PLS with structural data (1 condition, 1 group, n=80). While creating the Datamat, after the program loads all the 80 images, there appears an error message entitled "Too much data", with the content as:
ERROR: there is too much data to check zero variance, please delete some subjects in the "Edit Subject" window by clicking "Select Subjects" button in the session information window.
Would you kindly advise what was going on and how to solve this issue? Thank you very much in advance.
ATB,
Meng-chuan
Replies:
Untitled Post
I'm Online
Posted on 08/20/12 13:12:48
Number of posts: 229
There shouldn't be any problems with 80 datasets.
Can you provide some more information? What is the size of your masked datasets? What kind of structural data is being uploaded?
nancy
Can you provide some more information? What is the size of your masked datasets? What kind of structural data is being uploaded?
nancy
Untitled Post
I'm Online
Posted on 08/28/12 14:32:13
Number of posts: 291
I looked into the error message in my code. It seems that the data is too big to prevent the following simple command to be operated properly:
try
var(datamat);
catch
error message;
end
In order to further look into the problem, it would be better for me to duplicate exactly what you did. Would it be possible for you to upload your raw images, and tell me how you create the datamat?
Jimmy
try
var(datamat);
catch
error message;
end
In order to further look into the problem, it would be better for me to duplicate exactly what you did. Would it be possible for you to upload your raw images, and tell me how you create the datamat?
Jimmy
Untitled Post
I'm Online
Posted on 08/28/12 15:50:56
Number of posts: 229
I had no problems creating a datamat with 80 subjects, using *.nii files, each image is 256*256*160, and final datamat is 1516237 common elements per subject, from 1521554 elements in the brain mask.
I can run a behavioural analysis with no statistical tests... Running bootstraps and permutations would require something more than my poor laptop running Win7,64K, 4GB ram ;-)
nancy
.... so please do post the details of your image parameters, and if you can double check that your mask removes most of the background voxels, that would be a good thing to do as well. What is your operating system & memory?
I can run a behavioural analysis with no statistical tests... Running bootstraps and permutations would require something more than my poor laptop running Win7,64K, 4GB ram ;-)
nancy
.... so please do post the details of your image parameters, and if you can double check that your mask removes most of the background voxels, that would be a good thing to do as well. What is your operating system & memory?
Untitled Post
Posted on 08/30/12 10:41:30
Number of posts: 6
quote:
I had no problems creating a datamat with 80 subjects, using *.nii files, each image is 256*256*160, and final datamat is 1516237 common elements per subject, from 1521554 elements in the brain mask.
I can run a behavioural analysis with no statistical tests... Running bootstraps and permutations would require something more than my poor laptop running Win7,64K, 4GB ram ;-)
nancy
.... so please do post the details of your image parameters, and if you can double check that your mask removes most of the background voxels, that would be a good thing to do as well. What is your operating system & memory?
Dear Jimmy and Nancy,I had no problems creating a datamat with 80 subjects, using *.nii files, each image is 256*256*160, and final datamat is 1516237 common elements per subject, from 1521554 elements in the brain mask.
I can run a behavioural analysis with no statistical tests... Running bootstraps and permutations would require something more than my poor laptop running Win7,64K, 4GB ram ;-)
nancy
.... so please do post the details of your image parameters, and if you can double check that your mask removes most of the background voxels, that would be a good thing to do as well. What is your operating system & memory?
Thank you for checking this. I was working on a shared remote PC with a 32 bit system and 4GB RAM (but not sure how much is allocated to individual user) when this error happened. However when I then moved the data to another 64 bit machine with larger memory capacity there was no problem at all, and all the following procedures work smoothly. Therefore I guess this may be due to a memory allocation issue? -Meng-chuan
Login to reply to this topic.
