scrambled output from mask_data.m

[GoogleCodeExporter]

What steps will reproduce the problem?
1. load epi data using mask1 (epi input may have some zeros in it)
2. subj = mask_data(subj,'pattern','epi','mask2_subset_of_mask1')
3. view output epi using view_montage or other program

What is the expected output? What do you see instead?
Expected output: aligned masked data (input was smoothed ahead of time) 
Actual output: misaligned masked data (apparently correct spatial coordinates, 
but incorrect distribution of pattern based on adjacent jaggies. This was 
observed only in the upper half of the image; the bottom half was OK)

What version of the product are you using? On what operating system?
Rondo install of MVPA as of Feb 21 2011 on Matlab v.7.10.0

Please provide any additional information below.
In the process of extracting spatial coordinates, I noticed the feature 
selection output was looking garbled (in all subjects). I traced the problem 
back to mask_data.m, which produces filtered epis that look like the attached 
(misaligned upper brain, normal lower brain).

I'm not sure whether others will encounter this problem - it might be an 
artifact of my input data, which includes some zeros within the anatomical mask 
(for reasons I won't go into).

Original issue reported on code.google.com by jpopp...@gmail.com on 21 Feb 2011 at 9:52

[GoogleCodeExporter]

Which function are you using to load data into MVPA initially?

Original comment by gmcgrath...@gmail.com on 21 Feb 2011 at 10:07

[GoogleCodeExporter]

I'm using load_afni_pattern - I checked the input to mask_data, and it looks 
good.

Jordan

Original comment by jpopp...@gmail.com on 21 Feb 2011 at 10:10

[GoogleCodeExporter]

[deleted comment]

[GoogleCodeExporter]

I believe this resolved itself as self inflicted as mask_data isn't a function 
of the MVPA package (we would be more than happy to include it in the contrib 
area when it's kinks are worked out however!).  Will close for now, reopen this 
if it turns out to be an issue on our end.

Original comment by gmcgrath...@gmail.com on 24 Feb 2011 at 9:26

• Changed state: Invalid

[GoogleCodeExporter]

Self-inflicted is quite right. Sorry about the mix-up, and thanks for the quick 
help!

Original comment by jpopp...@gmail.com on 24 Feb 2011 at 9:31