Show all posts by user
Dear AFNI users-
We are very pleased to announce that the new AFNI Message Board framework is up! Please join us at:
https://discuss.afni.nimh.nih.gov
Existing user accounts have been migrated, so returning users can login by requesting a password reset. New users can create accounts, as well, through a standard account creation process. Please note that these setup emails might initially go to spam folders (esp. for NIH users!), so please check those locations in the beginning.
The current Message Board discussion threads have been migrated to the new framework. The current Message Board will remain visible, but read-only, for a little while.
Sincerely,
AFNI HQ
History of AFNI updates
Is http://afni.nimh.nih.gov/pub/dist/atlases/atlases_latest.tgz the canonical URL for getting the latest atlases? I've started building from source on a semi-regular basis (because for some reason it improves performance, and because I like the bleeding edge) and I want to be able to include updating the atlases in my build script.
by
Isaac Schwabacher
-
AFNI Message Board
If you really want to patch subprocess, use this:
--- /usr/lib64/python2.6/subprocess.py 2010-11-22 15:03:35.000000000 -0600
+++ /home/schwabacher/subprocess.py 2014-12-18 11:35:10.373428028 -0600
@@ -577,6 +577,8 @@
class Popen(object):
+ _shell_ = '/bin/sh'
+
def __init__(self, args, bufsize=0, executable=None,
stdin=None, stdout=None, stderr=N
by
Isaac Schwabacher
-
AFNI Message Board
It's a lowercase '-c'. Uppercase '-C' sets `noclobber`:
$ tcsh
% sh -C <<EOF
? echo first > temp.txt
? echo second > temp.txt
? cat temp.txt
? EOF
sh: line 2: temp.txt: cannot overwrite existing file
first
EDIT: Also, I would be *extremely* wary of changing the shell used by subprocess, especially if you're using the system python. It's
by
Isaac Schwabacher
-
AFNI Message Board
I tried this and it looks like all my environment variables are being inherited correctly:
$ tcsh
% setenv test val
% echo $test
val
% sh -c 'echo $test'
val
% ipython
In [1]: import os, afni_base
In [2]: os.environ['test']
Out[2]: 'val'
In [3]: com = afni_base.shell_com('echo $test')
In [4]: com.run()
val
Out[4]: 0
In [5]: com = af
by
Isaac Schwabacher
-
AFNI Message Board
My suspicion is that you've stumbled on one of the places where AFNI's python scripts don't play nicely with nifti datasets. Try converting your inputs to AFNI format and trying again.
by
Isaac Schwabacher
-
AFNI Message Board
Have you tried using the most recent OS X binaries? From what I've read on this board, it's pretty common for the binaries from one version of OS X to work on other versions.
by
Isaac Schwabacher
-
AFNI Message Board
Try setting the color bar to continuous ("**") and select a color scale instead.
by
Isaac Schwabacher
-
AFNI Message Board
Hmm... actually, alpha being turned on does not seem to be necessary for this. Steps to reproduce:
- open AFNI
- open a second viewer window
- overlay the same dataset in both viewers (the underlays can differ)
- turn on clusterize in [ B] (and set thresholds such that this causes some clusters to disappear)
- click the blue location slider in
- click the blue location slider in
-
by
Isaac Schwabacher
-
AFNI Message Board
You weren't kidding! Opening the same dataset in two different viewers, one with alpha on and the other with clusterize, has strange results when I click on the blue location slider in an image window.
by
Isaac Schwabacher
-
AFNI Message Board
Also, I notice that there's no backslash at the end of the line containing "-suffix _align", and that each subsequent input line begins with a '?'. Could that be part of your problem?
by
Isaac Schwabacher
-
AFNI Message Board
Would it be possible to make save an image that's overlaid over a white background with an alpha gradient perpendicular to the color gradient, as in the Allen, et al. paper?
by
Isaac Schwabacher
-
AFNI Message Board
I was thinking that you'd just have full transparency when thr = 0, full opacity when |thr| >= thr_slider, and linear scaling in between, since the threshold slider would be otherwise useless in this mode. So all you'd need would be an on/off switch, which could be just an extra value on the opacity control that already exists in the viewer window. (What do you call such a control?
by
Isaac Schwabacher
-
AFNI Message Board
Hi Emperor Zhark,
I'm rather taken with the visualizations in the paper Allen, et al. 2012 and it seems to me that they shouldn't be too hard to generate in AFNI. I know, however, that those are famous last words, and that the last time I ventured into the main AFNI viewer code, I emerged empty-handed. Is it feasible to have a setting to use the threshold image for the alpha channe
by
Isaac Schwabacher
-
AFNI Message Board
Your original problem is that you're not substituting commands in the way that you want to. If you're using bash, you can accomplish this with process substitution:
...
-stim_times 1 <(sed -n '1,1p' ./Onset_Sub"$subj"/Main1.txt) 'BLOCK4(0.8,1)' -stim_label 1 Event1 \
...
The "<(command)" syntax runs "command" in a separate p
by
Isaac Schwabacher
-
AFNI Message Board
The figure in that paper incorrectly suggests that the method used was an average of the correlations, which can be accomplished in the way you suggest, but buried in the text is a description that matches 3dTcorrMap -Zmean, which doesn't reduce to the easier case:
QuoteWeighted global brain connectivity (wGBC) analysis
See the middle left portion of Fig. 2 for an outline of wGBC analy
by
Isaac Schwabacher
-
AFNI Message Board
It's just saying that it's giving you back a dataset that tells you where the brain is (which you requested with -mask_vol) instead of giving back a dataset that's just the brain (which is what you'd get with -orig_vol). From the help:
-norm_vol: Output a masked and somewhat intensity normalized and
thresholded version of the input. This is the default
by
Isaac Schwabacher
-
AFNI Message Board
Think of it like this: all of the betas and statistics that 3dDeconvolve outputs comes out of a gltsym, but the syms for the stims you input are generated by default, without you having to request them. So if you have
3dDeconvolve ... \
-stim_times_AM2 1 foo.1D "BLOCK(4,1)" -stim_label 1 foo \
-stim_times_AM2 2 bar.1D "TENT(0,10,6)" -stim_label 2 bar \
...
the
by
Isaac Schwabacher
-
AFNI Message Board
I'm not sure what happened with that first one, but it looks like you don't have the GNU Scientific Library installed. This StackExchange thread will help.
by
Isaac Schwabacher
-
AFNI Message Board
What's the output of these commands: ldd "$(which 3dSkullStrip)"
ls /{,usr/{,local/}}lib*/libgsl*
by
Isaac Schwabacher
-
AFNI Message Board
Use a GLT:
3dDeconvolve \
-stim_times 1 foo CSPLINzero(0,20,11) \
\
-gltsym 'SYM +foo[3] -foo[2]' \
by
Isaac Schwabacher
-
AFNI Message Board
There's an AFNI program called SurfDist. Maybe it does what you want?
by
Isaac Schwabacher
-
AFNI Message Board
The reason I ask is that I put together an automated tool to download the source tarball whenever it changes and commit it to a git repository; for this usage, it's probably necessary to just unconditionally pull the tarball and then compare it.
EDIT: Fixed to just use the archived mtime of the newest member of afni_src.tgz.
by
Isaac Schwabacher
-
AFNI Message Board
There aren't any 2014.10 dates in http://afni.nimh.nih.gov/pub/dist/tgz/; does this mean that those dates aren't updated for minor changes? Or the packages available there aren't updated? The dates in pub/dist/bin are even older.
by
Isaac Schwabacher
-
AFNI Message Board
1200 volumes * 128 * 128 * 40 voxels / volume * 4 bytes / voxel ≈ 3 GB, so 8 GB isn't enough RAM to store the input, output and mask in physical memory together (even though the mask is stored at 1 byte / voxel on disk, 3dcalc loads it as 32-bit floats), assuming your datasets are at the same spatial resolution as ours.
FMRI datasets are big.
by
Isaac Schwabacher
-
AFNI Message Board
Looks like it wasn't `cd niml/` that time. But google suggests that -no-cpp-precomp is a clang-gcc option that tells clang-gcc to behave more like GNU-gcc. Try grepping for and excising it from the Makefile.
by
Isaac Schwabacher
-
AFNI Message Board
Aliases are only expanded in interactive shells; make's shell is probably finding the clang gcc. Try adding "CC=/usr/bin/gcc" to your Makefile.
EDIT: to test this, you can say this:\gcc --versionQuoting the command disables alias expansion.
EDIT 2, while trying to build AFNI myself: "SHELL=/usr/bin/gcc" is probably a bad idea...
by
Isaac Schwabacher
-
AFNI Message Board