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
I suspect this could be caused by a few things. 1) Make sure you're using the Lion binaries. 2) You shouldn't need fink anymore. Feel free to delete it if you're not using it for something else. 3) Be sure to set the DYLD_FALLBACK_LIBRARY_PATH.
See this post:
by
Peter Molfese
-
AFNI Message Board
Gaurav-
Our scanner also does not put the slice timing information into the DICOM and resulting NIFTI files (converted with either dcm2nii or Dimon/to3d). One solution is to add the following line to your afni_proc.py scripts. In our case, the slices are alternating starting at 1 instead of 0 - substitute whatever your slice sequence is for alt+z2.
-tshift_opts_ts -tpattern alt+z2
T
by
Peter Molfese
-
AFNI Message Board
3dvolreg is part of AFNI but does not use OpenMP. Some of the OpenMP enabled programs that you might check are: 3dAllineate, 3dREMLfit, and 3dClustSim.
The errors you are getting are interesting. Can you tell us more about your system setup? Do you have fink, MacPorts, or Homebrew installed? What's your system path setup?
by
Peter Molfese
-
AFNI Message Board
Odd that it would stop working suddenly. I wonder if a RedHat update broke something. Have you installed the GNU Scientific Library (GSL)? Or try reinstalling it?
64-bit machine
yum install gsl.x86_64
by
Peter Molfese
-
AFNI Message Board
Hi Daniel,
I have experience with BOINC, I think it's a great system. I don't think it would be my first choice for processing fMRI data. You could wrap your AFNI commands using the wrapper app. I've used it for single programs, but not for a series of programs called through a shell script (e.g. afni_proc.py). In terms of using computers all over campus, I don't think
by
Peter Molfese
-
AFNI Message Board
Let me start off by saying that Xgrid has been discontinued and when you upgrade your Mac Pros to 10.8 (Mountain Lion) or later, you won't be able to use it anymore. Apple has stated that they will not open source the technology so ultimately you may end up migrating to something like Hadoop (http://hortonworks.com), Condor (http://research.cs.wisc.edu/condor/), or GNU Parallel (http://www.
by
Peter Molfese
-
AFNI Message Board
The general case from the documentation is:
Example 1: Segmenting a skull-stripped T1 volume with:
Brain mask, No prior volumes, Uniform mixing fraction
3dSeg -anat anat.nii -mask AUTO \
-classes 'CSF ; GM ; WM' -bias_classes 'GM ; WM' \
-bias_fwhm 25 -mixfrac UNI -main_N 5 \
-b
by
Peter Molfese
-
AFNI Message Board
I use the DYLD_FALLBACK_LIBRARY_PATH
export DYLD_FALLBACK_LIBRARY_PATH=/opt/X11/lib:/usr/local/abin
I haven't had any problems lately. There is a pending Xquartz patch that I haven't applied yet.
by
Peter Molfese
-
AFNI Message Board
I think you could get a much better computer than the one you listed for that price. Consider the following options:
1. Build your own - often the cheapest route with a good money to performance ratio. I'd suggest newegg.com
2. Buy refurbished - this can be from the manufacturer (e.g. Apple: Apple Store refurbished) or some ebay sellers are authorized computer resellers - I have had gr
by
Peter Molfese
-
AFNI Message Board
Thanks for the tip. I just checked the dimensions of all of the input files and all of the output files (shown below), they seem to match. vbm_stats.nii.gz here is the output of the 3dttest++. Oddly if I put the stats file as the underlay and the aveBrain (3dMean of all datasets going into 3dttest++), the datasets look to be roughly the same size.
aveBrain.nii.gz 1.500000 -1.50
by
Peter Molfese
-
AFNI Message Board
Hi Cynthia,
I don't think I've ever run into a problem with an upper limit of specifying GLTs. How many are you attempting to run? Have you tried setting the -num_glt flag in 3dDeconvolve?
You could always try running 3dD with -xrestore. See Gang's helpful post here .
by
Peter Molfese
-
AFNI Message Board
I've run some structural data through the VBM8 (SPM8 toolbox) and then run into trouble with running 3dttest++ on the resulting files. I've used this procedure before (since I prefer to operate in AFNI land). Essentially when I load the overlay stats over a mean image (made with 3dMean), I end up with the stats image looking considerably larger than the underlay (see below). Running
by
Peter Molfese
-
AFNI Message Board
We have a Pegasus R6 Thunderbolt drive and Drobo Pro for nearly the same purpose. You can pair either with a Mac Mini (helpful for the Pegasus if you don't have a Thunderbolt equipped Mac) and add Mac OS X Server to do Time Machine backups over the network. Or use rsync.
by
Peter Molfese
-
AFNI Message Board
I saw in the recent change allowing @auto_tlrc with the option of -init_xform AUTO_CENTER to automatically run @Align_Centers on datasets. I wanted to try it out (because I end up running @Align_Centers quite a bit) so I added "-tlrc_opts_at -init_xform AUTO_CENTER" to my afni_proc.py script. On data where the centers are far apart, the option works perfectly. But on data where they
by
Peter Molfese
-
AFNI Message Board
1. See this thread for the info on the DYLD_FALLBACK
2. I haven't used tcsh in about 10 years, I'll defer to someone else on those specific questions.
3. You do not need to install Xcode to run AFNI... BUT I would really suggest you have it handy for users who want to do more than scripting commands together. If you're installing 10.7.x, I'd say Xcode 4.5. Don't f
by
Peter Molfese
-
AFNI Message Board
With 10.7 you won't need fink. Do the following to install afni in your home directory:
0. cd ~/
1. curl http:// afni.nimh.nih.gov/pub/dist/tgz/macosx_10.7_Intel_64.tgz > macosx_10.7_Intel_64.tgz
1a. REMOVE space between // and afni. I put it in to make it appear correctly on message board.
2. tar -xvf macosx_10.7_Intel_64.tgz
3. mv macosx_10.7_Intel_64 abin
4. nano .bashrc
4a.
by
Peter Molfese
-
AFNI Message Board
I have been experimenting with the nonlinear warping in AFNI. I noticed that when using 3dNwarpApply, the program does not currently update the space/view of a dataset to fit the warp. For example, if I am doing something like the code below, applying a nonlinear warp to a stats file, the output ends up being in +orig instead of +tlrc. I found that using 3drefit can solve the problem, but wasn
by
Peter Molfese
-
AFNI Message Board
I would recommend that you adjust the watershed values in freesurfer's skullstrip and then continue running the rest of the freesurfer recon-all pipeline. There are details on the Freesurfer wiki Fixing a Freesurfer Skull-strip.
Another option (though I've never tried it), is you can tell recon-all to not do skull-stripping using the -noskullstrip option. I believe you can then t
by
Peter Molfese
-
AFNI Message Board
From the error message, I am guessing that you don't have an X11 client installed. Unlike previous versions, Mac OS X 10.8 doesn't ship with an X11 client. You will need to install XQuartz in order to run AFNI.
Download XQuartz here
It is identical to the X11 application that you used to run in Mac OS X 10.5/10.6.
From the webpage description:
QuoteThe XQuartz project is
by
Peter Molfese
-
AFNI Message Board
I also haven't had any problems using it on a laptop mostly to view. I use Homebrew instead of fink or macports.
by
Peter Molfese
-
AFNI Message Board
Are the errors using freesurfer's pipeline or using 3dSkullStrip? What are you doing specifically?
by
Peter Molfese
-
AFNI Message Board
Bumping up to see if there's any updates on the 3dMEMA error with the "-cio" option.
by
Peter Molfese
-
AFNI Message Board
In the blur esimation block of afni_proc.py generated scripts, the script gets the number of TRs for each run.
example:
set tr_counts = ( 109 109 103 109 109 109 109 109 109 109 )
The tr_counts variable is set after the tcat step. Is there an easy way to do this in my own code? My older code used an odd combination of 3dinfo and AWK that I would like to get away from.
by
Peter Molfese
-
AFNI Message Board
Could be a number of factors, like number of runs and TRs within the runs. Do you have more details about your analyses and setup?
If you have a multiprocessor computer, you can speed up the 3dDeconvolve loop using the following options to spawn multiple threads. See example 7 for more details.
-regress_opts_3dD -jobs 4
by
Peter Molfese
-
AFNI Message Board
Hi Ziad-
Version info:
Version AFNI_2011_12_21_1014
[]
This is data in the volume, not on the surface. Surface analyses are on the list of things to do later on though. I used the gen_group_command.py to generate the 3dMEMA script.
input to gen_group_command.py:
gen_group_command.py -command 3dMEMA \
-dsets stats*.HEAD \
-prefix aud_word_mismatch \
-subs_betas 28 \
-subs_tst
by
Peter Molfese
-
AFNI Message Board
I am also running on 10.7.4 with R 2.15.1. Before I had found this thread to ignore the warning message, I also tried including the "-cio" option for 3dMEMA and got the following error. Any tips?
** Error:
Missing R_io.so
Error in if (an$type == "BRIK" && !is.na(an$view)) { :
missing value where TRUE/FALSE needed
Calls: process.MEMA.opts -> exists.
by
Peter Molfese
-
AFNI Message Board
I use 3dAttribute with the "-center" flag.
> 3dAttribute -center MPRAGEs015a001.nii.gz
-7.656250 -22.477341 -6.073975
by
Peter Molfese
-
AFNI Message Board
I was adapting (read: stumbling) through some really old scripts and found an alternative to specifying the filelist for anyone who is interested. The xargs command works great for long file lists as well.
cat myListOFiles.txt | xargs 3dbucket -prefix all_dsets
by
Peter Molfese
-
AFNI Message Board