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
That's odd, your output before said you were using bash. To add it to your tcsh:
echo "setenv DYLD_FALLBACK_LIBRAY_PATH ~otto/AFNI " >> .cshrc
by
Peter Molfese
-
AFNI Message Board
First run:
which afni
Then, making sure that you are in your home directory (the cd with nothing after it), take the result of "which afni" (something like ~/abin) and add it to your .bashrc (done here with echo and >>). I also link the bashrc and profile as some OS X programs automatically load one but not the other (or used to).
cd
echo "export DYLD_FALLBACK
by
Peter Molfese
-
AFNI Message Board
Have you tried running:
suma -update_env
The one time I had color issues, the above command helped.
by
Peter Molfese
-
AFNI Message Board
I get this message when the DYLD_FALLBACK_LIBRARY_PATH isn't set correctly. What is the output of
echo $SHELL
and
echo $DYLD_FALLBACK_LIBRARY_PATH
and
afni -version
by
Peter Molfese
-
AFNI Message Board
For most things, you just need to install XQuartz and set your DYLD_FALLBACK_LIBRARY_PATH to point to your AFNI installation. After that, most things should "just work".
If you want to use the uber_subject.py and uber_ttest.py programs, you will need to install some other libraries, the easiest way I have found to do this is:
install Xcode install homebrew brew install pyqt
by
Peter Molfese
-
AFNI Message Board
I am completely stumped. If it were me, I would try reinstalling the OS, skipping fink and MacPorts and just trying to get it running with Xcode installed.
by
Peter Molfese
-
AFNI Message Board
I figured out my problem. When 3dANOVA3 is running, it's generating files that are BRIK.gz format (thanks to the pigz compressor). After the 3dbucket and 3drefit stages complete successfully - 3dANOVA deletes all of the .HEAD files correctly but doesn't seem to figure out to delete the .BRIK.gz files.
The older install of AFNI wasn't writing .BRIK.gz files, so the program di
by
Peter Molfese
-
AFNI Message Board
That is really puzzling. Also odd that the libraries in /usr/lib that are referenced I don't see on my systems here.
Perhaps with more information:
output of:
which gcc
gcc -v
xcode-select -version
This may be approaching the dreadful "Reinstall the Operating System" stage...
by
Peter Molfese
-
AFNI Message Board
Can you paste the contents of your .cshrc and the updated echo $PATH?
by
Peter Molfese
-
AFNI Message Board
Isabelle,
You might try removing the path to fink (/sw) and MacPorts (/opt/local) from your .login and .cshrc and see if that fixes your problem. Once edited, you will need to open a new terminal window. Check that all references to fink and MacPorts are gone and then try your align_epi_anat.py script again. If this works, then you could try to change the way that you add the fink and MacP
by
Peter Molfese
-
AFNI Message Board
I'm seeing a similar behavior with the July 2, 2013 binaries. 3dANOVA3 runs and the 3dbucket command generates a final output image containing all of the contrasts, but it looks like 3dANOVA3 isn't deleting the temporary files associated with each main effect, interaction, or contrast. I tried running on an older version of AFNI (Apr 23, 2013), and found that 3dANOVA3 correctly remove
by
Peter Molfese
-
AFNI Message Board
Isa,
I suspect that you're getting some interference between fink, macports, and the default xcode installation. In addition to the path and shell info, what are the outputs of:
port installed
fink --installed
It's generally not recommended to have both installed since they can conflict with one another!
by
Peter Molfese
-
AFNI Message Board
Before switching binaries:
1) Do you have fink or MacPorts installed?
2) What is the output of echo $SHELL? and echo $PATH?
by
Peter Molfese
-
AFNI Message Board
I use the "Save Layout" feature under "Define Datamode -->" and "Misc" to do this.
by
Peter Molfese
-
AFNI Message Board
Durai-
A member of the AFNI team may have a better answer... I ran the 3dvolreg command on about 30 DTI scans that I had "around", but I do not see an intensity change (to that level). It's possible that it's related to a previous discussed topic on Fourier interpolation having a difficult time with spikes in the data or other distortions. It could be that your particular
by
Peter Molfese
-
AFNI Message Board
Because you don't need fink for most things using the Mac OS X 10.7 binaries (which work on 10.8), I've been telling people to not use fink and instead install homebrew to get the necessary libraries. All you need is Xcode to begin.
1. Install Homebrew: ruby -e “$(curl -fsSL )”
2. Install PyQT: brew install pyqt
3. Set Python Path: export PYTHONPATH=/usr/local/lib/python2.7/site
by
Peter Molfese
-
AFNI Message Board
Hi Durai-
Can you list the commands you used for 3dvolreg and mcflirt?
Best,
Peter
by
Peter Molfese
-
AFNI Message Board
I'm not 100% certain on the implementation of 3dMVM.
But the short answer for statistics of whether to use a series of t-tests vs. ANOVA framework comes down to "it depends" (Maxwell & Delaney, 2004, p. 165). If you are primarily interested in simple paired contrasts between groups, and feel confident that the assumption of homogeneity-of-variance is true (all three groups
by
Peter Molfese
-
AFNI Message Board
What you are labeling the t-test in 3dMVM is really a GLT. The degrees of freedom for the contrast (GLT) would be the same as the full model (in this case 100).
by
Peter Molfese
-
AFNI Message Board
Can you tell us more about the error? Which tool were you using? unpacksdcmdir? This might be better on the freesurfer mailing list.
You might try converting the DICOM files to NIFTI via AFNI tools (e.g. Dimon) using a command similar to:
Dimon -infile_prefix path/to/dicoms/MR.* -dicom_org -GERT_Reco -quit
Alternatively, you could use dcm2nii or MRIConvert.
by
Peter Molfese
-
AFNI Message Board
Hi Emily,
You can use the -tlrc_base option in afni_proc.py to change the template that data are warped to. There are a few options:
MNI_avg152T1+tlrc.HEAD
MNI_caez_N27+tlrc.HEAD
More info on spaces is available on Daniel Glen's page here.
by
Peter Molfese
-
AFNI Message Board
Jessica-
To your original question, my quick test suggests that the implementation of 3dRegAna should give you almost identical results for Coef #3 and the F-stat, with the exception that the t-stat sub-brik will give you directionality, whereas the F-stat will display all differences as positive. I believe that 3dRegAna F-test evaluates if the Full Model is significantly different from the
by
Peter Molfese
-
AFNI Message Board
I have had some success copying/modifying an afni_proc.py script to grab the 1) pre-processed files from one "results" directory; 2) the motion estimates (from 3dvolreg); 3) the anatomic; and 4) the stim files. You could then use the modified afni_proc.py script to quickly generate a script doing just the masking (I think this was necessary to get 3dClustSim to run) and regression. Ex
by
Peter Molfese
-
AFNI Message Board
In your afni_proc.py command, having both a "-blocks" and a "-do_block" is somewhat repetitive (and might break something). You should be able to cutout the "-blocks tshift volreg blur mask scale regress" and get the same result that you are aiming for. I wonder if having both specified is giving afni_proc.py a hard time.
Is this error happening on other subje
by
Peter Molfese
-
AFNI Message Board
Hi Sarah-
Is this error for more than one participant? My first guess would be that there is something wrong with run 4 of that particular subject. If you run the following command inside the "results" folder created by afni_proc.py, what do you see?
3dinfo -prefix -d3 -n4 pb01*.HEAD
Also can you post the entire afni_proc.py command that you are using?
by
Peter Molfese
-
AFNI Message Board
That's an interesting one - sometimes happens if you quit apt-get or dpkg before they complete. Usually rebooting fixes it (so try rebooting first!).
But if that doesn't work, try this page.
Of course if you want to just throw caution to the wind, you can always do:
sudo rm /var/lib/dpkg/lock
sudo dpkg --configure -a
sudo apt-get update
by
Peter Molfese
-
AFNI Message Board
Try installing gsl with the following command:
sudo apt-get install gsl-bin
by
Peter Molfese
-
AFNI Message Board
Hi Sarah,
It looks like you don't have libgsl (part of the GNU Scientific Library). What OS (Mac or Linux) are you running?
This thread may be helpful:
by
Peter Molfese
-
AFNI Message Board
I haven't had any problems getting Ubuntu to dual boot on a PC following these instructions. You can also still buy some Windows laptops with Windows 7 installed.
I'm personally a Mac person and I run quite a few analyses in AFNI and Freesurfer on my Macbook Air (which is now 2 years old). You can run AFNI on somewhat modest (by today's standards) processors, so any of the Ma
by
Peter Molfese
-
AFNI Message Board
During the minutes it took me to write a response Rick had answered the question!
by
Peter Molfese
-
AFNI Message Board