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 3901 - 3930 of 4524
HI Giuseppe,
Thanks to Daniel for looking up the old thread.
What does nifti_tool say about your scl_* fields?
nifti_tool -disp_hdr -infile DATA.nii | grep scl_
- rick
by
rick reynolds
-
AFNI Message Board
Hi Susie, Hi Jackie,
The next bootcamp is tentatively scheduled for the week
of December 2nd. Don't buy any plane tickets yet, but that
week is very likely when it will be.
- rick
by
rick reynolds
-
AFNI Message Board
Hi Emily,
If you do not censor and you do not bandpass in afni_proc.py, then passing
the errts dataset would be fine. In that case you would not need any -ort
option (since those terms would have already been removed).
Alternatively, you could still censor in afni_proc.py, depending on what
you want to do. In addition to X.xmat.1D (with censoring), the AP script
would produce X.noce
by
rick reynolds
-
AFNI Message Board
Hi Gaurav,
Assuming your tents are TR-locked, yes. Either way, they are
indeed the betas estimated on the TENT grid. If the 3dDeconvolve
input was in percentage units, then the betas would be as well.
So yes, you have it right.
- rick
by
rick reynolds
-
AFNI Message Board
Hi Zhihao,
As with any of the pull down menus, you can right-click
just to the left of the menu to get a scrollable chooser.
In the case of atlas labels in the drawing plugin, there is
no actual text there. But you can still do it.
- rick
by
rick reynolds
-
AFNI Message Board
Hi Isabelle,
The run order for afni_proc.py is exactly as you give it
to the program. So it is up to you to provide the EPI
datasets in the order that they should be applied.
If you use separate 3dDeconvolve commands, then the
timing files would each have only the 2 rows (for 2 runs).
But if you give it all to afni_proc.py, then the files should
have lines for all 4 runs (since they w
by
rick reynolds
-
AFNI Message Board
Hi Isa,
Are you running afni_proc.py through regression? If so, then I
don't really see what you would gain from naming the EPI datasets.
The labels for the beta weights are what one typically cares for.
Otherwise, would you provide some more details about exactly
what you are hoping to do? Are you just using afni_proc.py for
preprocessing, and then running your own pair of 3dDec
by
rick reynolds
-
AFNI Message Board
Hi Emily,
Given Paul's suggestion of:
3dRSFC -nodetrend -mask MASKNAME -bp_at_end -prefix PREFIX 0.01 0.1 INPUT_NAME
and Ziad's comment that 3dRSFC would want to remove the same
model as 3dD, except for the bandpass regressors, it makes sense to
give 3dRSFC the input to 3dDeconvolve, plus a matrix of regressors
to orthogonalize to. Also, censoring would presumably not be done.
by
rick reynolds
-
AFNI Message Board
Hi Anders,
"Regress motion derivatives" means, include the temporal
derivatives of the motions parameters in the regression model.
Either way, the ordinary motion parameters are included.
- rick
by
rick reynolds
-
AFNI Message Board
Hi Rasmus,
According to Gang, those values are not properly normalized in
order to compare across different models, and what you suggest
is surely part of that. So that is where we start from (saying that
you should not do things like alter the number of stimuli or the
durations). It is generally used to optimize within a given design,
not across designs.
Note that MRT.py does not wo
by
rick reynolds
-
AFNI Message Board
Hi Nicole,
There seems to be a space between './' and the dataset prefix.
I think Daniel made some directory updates on 3/12, and your
version is from 3/11. Try updating one more time.
Thanks for pointing this out in any case.
- rick
by
rick reynolds
-
AFNI Message Board
Hi Isa,
Since your EPIs are across sessions, you should probably
align them within a session, and then each session to the
same anatomy. This is not optimal though.
Given that, it might be good to use @auto_tlrc and then
align_epi_anat.py yourself, and then have afni_proc.py
finish it off.
Otherwise, you could take the proc script, add a second
align_epi_anat.py command (for the s
by
rick reynolds
-
AFNI Message Board
Hi Nicole,
Since it sounds like you are using afni_proc.py, would you send
the output.proc file via email? Do you have that file (containing
all of the text output from the processing)? It might be helpful,
since this error is probably the result from an earlier one that is
not shown.
I'll PM you with my email address, in case you can send it.
- rick
by
rick reynolds
-
AFNI Message Board
Hi Mika,
Colm has a good suggestion, but I expect that you are
simply running out of RAM. I cannot tell exactly which
version of AFNI you are running, but it is probably
compiled to handle large datasets (even if on a 32-bit
system).
If you are able to open this in another program, then maybe
you just have some other programs currently running that
are taking up too much memory.
by
rick reynolds
-
AFNI Message Board
Hi Heather,
What are the 2 lines output by "afni -ver"?
Do you have my email address? If so, would you send
me the output after added "-debug 3" to the command?
- rick
by
rick reynolds
-
AFNI Message Board
You can also use "timing_tool.py -add_offset ..." to adjust the
timing files. See the -help output for details, including example 2.
- rick
by
rick reynolds
-
AFNI Message Board
Hi Jenny,
Scaling does not affect the correlations, so that step is unnecessary.
- rick
by
rick reynolds
-
AFNI Message Board
Hi Noah,
In addition to Nick's comments, note that -automask is
not appropriate on a residual time series, as you are
using here:
> 3dFWHMx -automask -input errts.1080+tlrc
Pass a proper EPI mask dataset instead.
You should also consider breaking runs apart, maybe
following what afni_proc.py does.
- rick
by
rick reynolds
-
AFNI Message Board
Hi Laurel,
There is no direct NN=1 value for the renderer. As you have seen,
the rmm (radius in mm) value controls the cluster connections.
If the voxels are isotropic, then rmm=voxel size should give the
same results as NN=1. However if the voxels are 2x2x4, say,
then the render cannot do NN=1 (for example, rmm=4 to get the
slice direction results in not only within slice diagonals,
by
rick reynolds
-
AFNI Message Board
Hi Gaurav,
You can use whatever tool (e.g. 3dROIstats, 3dclust) to make a
text file of the maximum cluster coordinates, and then use 3dUndump
with the -master, -xyz, -srad and maybe -orient options to create a new
dataset of the spheres.
The -orient option is a good idea, to tell 3dUndump what the x, y and z
coordinates mean, since they may not match those of the -master
dataset.
-
by
rick reynolds
-
AFNI Message Board
Hi Isa,
If you were using afni_proc.py for your preprocessing (which
we suggest), then you could apply the timing pattern via an
option like:
-tshift_opts_ts -tpattern alt+z
Note that 3dTcat also has new -tr and -tpattern options,
though they will only be available once we do a build (maybe
tonight).
- rick
by
rick reynolds
-
AFNI Message Board
Note that 3dTcat now has -tr and -tpattern options.
- rick
by
rick reynolds
-
AFNI Message Board
Hi AW,
Those numbers look good. To make yourself more confident,
compare the censored X-matrices to the uncensored one. They
should be identical (except for being restricted to the given TRs).
You can use either the global or local format for that option.
- rick
by
rick reynolds
-
AFNI Message Board
Hi AW,
There seems to be a problem with -nlast, perhaps a mix of whether
it applies to each run or all of them.
Go with -CENSORTR for now. Something like 3:0-200 should work,
assuming you have no more than 200 TRs in run 3. I'll try to take a
closer look...
- rick
by
rick reynolds
-
AFNI Message Board
Hi Emily,
It sounds to me like you want to examine connectivity maps
using an ANOVA, i.e the differences are presumably going to
come from a seed-based correlation analysis. Does that seem
correct?
So you would choose a seed location (I don't see how you plan
to do this), create normalized correlation values, and then
run an ANOVA on the results.
Otherwise, exactly what would
by
rick reynolds
-
AFNI Message Board
Hi Lisa,
For creating unit vectors of the time series, you might consider
what @compute_gcor does:
3dTnorm -polort 0 -prefix OUTPUT INPUT
The default scaling is the L2 norm, setting the sum of squares
equal to 1. If instead you want the standard deviation, then there
is a factor of sqrt(NT-1) to scale in. Maybe I should consider
adding such an option...
- rick
by
rick reynolds
-
AFNI Message Board
Hi Russell,
You seem to be describing a surface dataset, along with its normals
and curvature values.
If you are not comfortable with what is generated by FreeSurfer or
Caret and want to write you own, that might be a noble effort. But
you will almost certainly not want your data in ascii format. Anatomical
datasets typically have millions of voxels, and maybe a million just of
gray
by
rick reynolds
-
AFNI Message Board
Hi Lisa,
It is not clear to me what you mean by z-normalizing a time series.
Would you clarify how that should be done? Perhaps that means
scaling to a constant standard deviation?
And similarly, how would one scale betas to z-scores? Do you
mean applying the Fisher z-transform to R values? If so, that is
fairly simple. But R values are generally not called betas, so I
suspect you
by
rick reynolds
-
AFNI Message Board
Hi Sammy,
What version did you download? It must be 10.6 or even earlier.
The older versions required a handful of fink libraries.
Also, do you have fink installed?
- rick
by
rick reynolds
-
AFNI Message Board
You have spaces after the '\' characters on each of the
proceeding lines (given the "command not found" errors).
You can check your script with:
file_tool -test -infile 3dLME.sh
- rick
by
rick reynolds
-
AFNI Message Board