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
Results 4021 - 4050 of 4524
That "0 1" for bandpassing looks very suspicious!
Check those parameters. I guess there are no bandpass
regressors (assuming your TR is less than 5 seconds).
- rick
by
rick reynolds
-
AFNI Message Board
Are you doing steps 3 and 4? When uber_subject.py displays the
afni_proc.py command, edit that command removing -volreg_tlrc_warp,
then use ctrl-S (or file->save) to save it.
Then proceed with the analysis.
- rick
by
rick reynolds
-
AFNI Message Board
Hi Ana,
What are the 2 lines of output from "afni -ver"?
Maybe you are using 10.6.no.glib, which is no
longer good enough as it stands (because of the
change to 3dSkullStrip).
- rick
by
rick reynolds
-
AFNI Message Board
Okay, a fix is in. Let me know if you want a
distribution update for that sooner than tonight.
Thanks again,
- rick
by
rick reynolds
-
AFNI Message Board
Hey PT,
Yes, though apparently I forgot to remove an AP option in that case, so
you will need to make one edit.
0. Set all of your desired options.
1. In the "analysis initialization" section, remove the 'tlrc' processing block.
2. Use the left action button to create the afni_proc.py command (as usual).
3. Edit the afni_proc.py command, removing the '-volreg_tl
by
rick reynolds
-
AFNI Message Board
Hi Ana,
If you have 3dSkullStrip in the current directory and want to run it,
use: ./3dSkullStrip ...
But this thread is over a year old now, so maybe you are starting
from a different point. What OS X version are you using? If it is
at least 10.7, then you should try the 10.7 binaries and set
DYLD_FALLBACK_LIBRARY_PATH to $HOME/abin.
There was a recent change to 3dSkullStrip to d
by
rick reynolds
-
AFNI Message Board
Hi Peter,
Actually, I believe that when there is an odd number of
slices you will get alt+z, but for even, alt+z2.
I cannot speak for any of Siemen's craziness...
- rick
by
rick reynolds
-
AFNI Message Board
Hi Colm,
Peter pointed out the important part of using a packaged
that comes with shared libraries. On the mac you should
set DYLD_FALLBACK_LIBRARY_PATH to the location of
the binaries (and shared libraries).
- rick
by
rick reynolds
-
AFNI Message Board
I had forgotten about that, thanks for the reminder!
- rick
by
rick reynolds
-
AFNI Message Board
A fix for this is in, and will be available the next time
binaries are compiled. Since you care about compiling
from the source, the afni_src.tgz package has been
updated.
"afni_history 5" (for example) should mention this.
Thanks again for bringing it up,
- rick
by
rick reynolds
-
AFNI Message Board
Hi Peter,
We may actually be heading more in that direction in the future,
not only since fink cannot keep up with the frequent OS X updates,
but also to reduce the size of our binary set. We may distribute
shared libraries, regardless of OS type.
That may pose a challenge for programs that rely on PyQt4, but it
is a bridge to cross later.
-rick
by
rick reynolds
-
AFNI Message Board
Hi Kelvin,
Yes, that is a problem with 3dSkullStrip reading NIfTI datasets.
I will fix that and get back to you.
Thanks for bringing it up,
- rick
by
rick reynolds
-
AFNI Message Board
Hi Kelvin,
That is not a good sample dataset. It is a simple byte
format and does not have a normal brain to be extracted.
Try the anat+orig.HEAD and .BRIK from here:
AFNI_data6/afni
- rick
by
rick reynolds
-
AFNI Message Board
Hi Paula,
Just to offer a slightly different take on Daniel's commands,
note that 3dTstat has a -tdiff option, which is to take the first
differences. So to get the sum of squared differences:
3dTstat -tdiff -sos -prefix sos INPUT+orig
Then use 3dcalc to divide by NT - 1:
@ ntm1 = `3dinfo -nt INPUT+orig` - 1
3dcalc -a sos+orig -expr a/$ntm1 -prefix mean.sos.diffs
-
by
rick reynolds
-
AFNI Message Board
Are you looking at datasets that were created many years
ago (or with the old software), or are you looking at datasets
created with the current binaries?
It is possible that it is only the warning that is new.
Where the bucket datasets saved from a single execution
of 3dDeconvolve, or via 3dDeconvolve -xrestore, or maybe
some other way?
Since we don't actually know how you crea
by
rick reynolds
-
AFNI Message Board
Hi Catherine,
Maybe you simply need to type "rehash" (or open a new
terminal). What is the output from these commands?
which afni
afni -ver
echo $0
- rick
by
rick reynolds
-
AFNI Message Board
Hi Chris,
Use "3dinfo -VERB" to show the timing pattern (before all of the
sub-brick info).
- rick
by
rick reynolds
-
AFNI Message Board
Hi Gulhan,
Assuming the datasets had correct (if different) orientations
to begin with, then 3drefit is not an appropriate program to
use to change the orientation, 3dresample is.
What 3dresample does: changes the orientation on disk
to convert from one proper orientation to another.
What 3drefit does: alters the header, saying "ignore the
current orientation, it should really b
by
rick reynolds
-
AFNI Message Board
Hi Kristina,
It is just there as an additional option.
Some people use only motion derivatives, some only motion
parameters, some use both. The option was added to allow
capturing more of the motion effect in the case of resting state
analysis, but it does not necessarily produce a big gain. It is
less of a question in task-based FMRI.
In any case, we do not say that everyone shoul
by
rick reynolds
-
AFNI Message Board
Hi Frank,
From what I recall, you can get either DICOM or Par Rec formated
files from a Bruker scanner (is NIfTI an option?).
You can work with any of these, though NIfTI should be the easiest,
then DICOM, then Par Rec.
What format are you working with? Do you have the option to get
NIfTI or DICOM?
- rick
by
rick reynolds
-
AFNI Message Board
Hi Chris,
Old slice information is lost upon resampling since the old
slices generally do not exist, so slice timing (and obliquity
information) are thrown out.
Your case is a little more simple though, since you are just
reorienting the original slices (the slice pattern and locations
did not change). So I guess you might use 3drefit to copy the
needed time info from the original dat
by
rick reynolds
-
AFNI Message Board
Hi Oori,
That can be found in the "Display Options" menu of any
image window (click on 'Disp' in the lower left corner).
What you describe is a projection +/- 5 slabs.
a. click 'Disp' in the z-plane image window
b. set 'Project' to Maximum (or whatever)
c. set 'Slab' to 5 (so +/- 5 slices)
You mentioned doing this to functional data,
by
rick reynolds
-
AFNI Message Board
Hi Valur,
InstaCorr goes through a setup phase to convert a dataset into
something it can process quickly. While it could surely be applied
to a dataset created by the realtime plugin, it is not set up to
process a dataset as it is acquired.
- rick
by
rick reynolds
-
AFNI Message Board
Hmmm, you have commented out the CCFAST line that included -m64
and uncommented the line that has -march=i686. But you do not have
an i686 architecture.
Consider starting with the openmp_64 Makefile, instead.
- rick
by
rick reynolds
-
AFNI Message Board
It looks like frappr.com is perhaps no longer active.
- rick
by
rick reynolds
-
AFNI Message Board
For particular programs, we really just have whatever regression
tests are used internally (and individually). Some programs have
such tests, some do not. It is up to each author to worry about
what they are doing.
If there are specific things on your mind, I would be happy to ponder
them over email.
- rick
by
rick reynolds
-
AFNI Message Board
There is no public archive, except that we do have an
internal one (partial, though).
Please contact me via either email or PM to discuss
what you might be interested in.
- rick
by
rick reynolds
-
AFNI Message Board
Hi JQ.Dai,
Consider using clustering (maybe defined via slow_surf_clustsim.py).
However, if the analysis was not done on the surface, then clustering
in the volume first would make more sense (since the surface would
simply be for display in that case).
- rick
by
rick reynolds
-
AFNI Message Board
Hi Mahen,
The mean of baseline-removed signal would typically not be
exactly zero but close to it (it would be the mean of the
signal of interest, which might be 1 or 2 for a very active
voxel, say). In any case, such a scaling would surely not
do anything helpful.
The trend does not really affect the scaling since Legendre
polynomials are used to model the baseline, and they are all
by
rick reynolds
-
AFNI Message Board
Hi Peter,
Yes, remove the first 5 TRs from each file, then concatenate
them.
To delete the first 5 rows from a bunch of motion files, use
1dcat or 1d_tool.py or head or tail. Maybe the former 2 would
be easier, since they do not require knowing the total number
of TRs.
Consider:
foreach mot_file ( motion*.1D )
1dcat $mot_file'{5..$}' > short.$mot_file
end
cat
by
rick reynolds
-
AFNI Message Board