write a statistic_image to file about canlabcore HOT 7 CLOSED

The 'mni' option in the write function involves resample_space, which might cause the error for statistic_image. I committed a small fix on that 8/4/15. Could you try this again and see if there is still an error?

from canlabcore.

The ‘mni’ flag was sort of a kludgy hack to get the FOV in a space for other viewers. It never worked for statistics_image() writes. Thanks Wani for fixing that. Luka and I added that awhile ago and that is something I was always meaning to fix.

On Aug 5, 2015, at 12:22 PM, Choong-Wan Woo [email protected] wrote:

The 'mni' option in the write function involves resample_space, which might cause the error for statistic_image. I committed a small fix on that 8/4/15. Could you try this again and see if there is still an error?

—
Reply to this email directly or view it on GitHub #3 (comment).

from canlabcore.

I didn't do it. Tor made a change in "resample_space" to make it work for statistic_image, but there was a small problem, which I fixed. Hopefully the 'mni' option works for statistic_image for now. I haven't tested it. Yoni will let us know and close the issue, if it works.

from canlabcore.

No, code still crashes when writing statistic_image with 'mni' flag. Here
is the error:

write(t, 'fname', 'temp.nii', 'mni')
Index exceeds matrix dimensions.

Error in image_vector/resample_space (line 88)
if ~isempty(obj.p), obj_out.p = p(Vto.wh_inmask, :); end

Error in image_vector/write (line 59)
obj = resample_space(obj,mni);

On Wed, Aug 5, 2015 at 11:23 AM, Choong-Wan Woo [email protected]
wrote:

I didn't do it. Tor made a change in "resample_space" to make it work for
statistic_image, but there was a small problem, which I fixed. Hopefully
the 'mni' option works for statistic_image for now. I haven't tested it.
Yoni will let us know and close the issue, if it works.

—
Reply to this email directly or view it on GitHub
#3 (comment).

from canlabcore.

It's now working. I fixed it by converting statistic_image into fmri_data before using resample_space.

from canlabcore.

hmmm, so resample_space is broken for statistic_image?

On Aug 5, 2015, at 1:24 PM, Choong-Wan Woo [email protected] wrote:

It's now working. I fixed it by converting statistic_image into fmri_data before using resample_space.

—
Reply to this email directly or view it on GitHub #3 (comment).

from canlabcore.

Actually, you added a comment saying that "this may break if nvox (total in
image) is different for 2 images..." in line 67 a while ago, and the
function exactly break there because of that exact reason (in Yoni's case).

On Wed, Aug 5, 2015 at 4:27 PM, Tor Wager [email protected] wrote:

hmmm, so resample_space is broken for statistic_image?

On Aug 5, 2015, at 1:24 PM, Choong-Wan Woo [email protected]
wrote:

It's now working. I fixed it by converting statistic_image into
fmri_data before using resample_space.

—
Reply to this email directly or view it on GitHub <
https://github.com/canlab/CanlabCore/issues/3#issuecomment-128118513>.

—
Reply to this email directly or view it on GitHub
#3 (comment).

Choong-Wan (Wani) Woo
Web http://wanirepo.github.io/ | Code https://github.com/wanirepo/ | Google
Scholar https://scholar.google.com/citations?user=fZLY5H8AAAAJ&hl=en&oi=ao
| Lab http://wagerlab.colorado.edu/
PhD candidate
Cognitive and Affective Neuroscience Lab
University of Colorado Boulder

from canlabcore.