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
Time starts at 0, so the last point in your dataset is at (524-1)*0.7 = 366.1 seconds. That is, the nth time point, for n=1,2, ..., occurs at time (n-1)*TR.
by
Emperor Zhark
-
AFNI Message Board
It uses the global average, not the per-run average -- sorry.
by
Emperor Zhark
-
AFNI Message Board
The closest analogy to FSL that I would recommend in AFNI would be to eschew the polynomial regressors in AFNI and instead put in trigonometric regressors to pick up the low frequencies you want filtered out, and let the linear regression + pre-whitening model do it all at once." So by eschewing polynomial regressors you mean to change the polort option to 0, correct? But I do not understand
by
Emperor Zhark
-
AFNI Message Board
I'm afraid you are doing things we don't recommend; also, we don't use FSL (ever) so we can't comment on trying to make AFNI match what FSL (or SPM) does -- I wouldn't even try. Both of these situations make it difficult to advise you.
I will mention that the combination of high-passing and using polynomial baseline regressors in AFNI is not a reasonable idea -- the n
by
Emperor Zhark
-
AFNI Message Board
Methods development in FMRI (which mean whatever you want it to mean):
Peter's lab space is right next to Emperor Zhark's, by the way.
by
Emperor Zhark
-
AFNI Message Board
Self-healing users == the best kind. Congratulations on solving your own problem!
by
Emperor Zhark
-
AFNI Message Board
Is your AFNI up to date? That is, does the command
afni -ver
show a compile date in the last month or so?
by
Emperor Zhark
-
AFNI Message Board
The duration and the stimulus times (in the -stim_times files) are both specified in seconds. If you are getting too much overlap, then you are either giving incorrect inputs to the program, or you have a mismatch between your experiment design and your expectations.
How are you diagnosing "extreme overlap of my blocks"? I personally would do it by graphing the relevant column(s) f
by
Emperor Zhark
-
AFNI Message Board
This opening is in a group that funds instruments for NIH-funded investigators. US Citizenship is a requirement, as is experience with at least one major class of instruments. Applications are due by Wed, 19 Aug 2015.
Link for people who are already Federal employees (or used to be):
Link for people who are not Federal employees:
by
Emperor Zhark
-
AFNI Message Board
3dQwarp is tricky to use on people with "abnormal" brains -- people with resections, tumors, significant dysplasia, .... In some cases, probably nothing useful can be done -- see the image in this Wikipedia article, for example:
In other cases, the "-emask" option might make it possible to get a good alignment using the "normal" part of the brain. Infant brains
by
Emperor Zhark
-
AFNI Message Board
I wrote 3dclust back in the Paleolithic days of AFNI, and my decision then to report only 1 decimal place was for 2 reasons:
Keeping the table compact Not pretending to have more precision than really exists
About the latter point: even printing to 0.1 mm accuracy is really too much -- considering FMRI voxels are usually 2-3 mm in size. I can't see how it's reasonable to expect pre
by
Emperor Zhark
-
AFNI Message Board
Hmmm -- I can't get this to happen here, on an ssh connection where X11 isn't forwarded (e.g., neither afni nor suma will start).
I do notice you are somewhat out-of-date -- for example, your align_epi_anat version is 1.48, and we are now at 1.50. Update your afni binaries and try again. Normally, the command
@update.afni.binaries -d
would do the job, but since you seem to be
by
Emperor Zhark
-
AFNI Message Board
That's Emperor Zhark just added it to the documentation.
I'm impressed by how fast you noticed that.
by
Emperor Zhark
-
AFNI Message Board
That's clever, Dominic and Gang. I wrote 3dttest++ and never thought of using the covariates for a 3-way interaction analysis. Since the t-statistics for the slopes of the inputs vs. the covariates are marginal, they factor out the linear effects of the covariates on the input values.
by
Emperor Zhark
-
AFNI Message Board
I prefer the 1996 paper be cited (at least), out of date as it is, because that makes it easy for me to track how much AFNI is used. I'm doing this now, for example, since I have to justify my continued funding very soon, so being able to say that AFNI was cited 1601 times in the last 4 years is (hopefully) impressive.
by
Emperor Zhark
-
AFNI Message Board
That was a mistake.
I just committed a change to InstaCalc to allow you to do this interactively as well -- instead of "Dataset: Value", choose "Dataset: Pvalu" from the menu item just to the right of the symbol name ("A", "B", etc.). And then choose the sub-brick index that you want converted on-the-fly. Presumably the expression would just be the let
by
Emperor Zhark
-
AFNI Message Board
I'll add that you can use the Histogram or Histogram:Multi plugin to interactively graph the p-values, once they have been converted. Or of course use 3dhistog and then 1dplot.
by
Emperor Zhark
-
AFNI Message Board
All the functions already existed (obviously). There are only 96 lines of code in the program, and most of those are simple re-writes of code fragments from other programs (e.g., opening datasets, parsing the "-prefix" option, etc.).
by
Emperor Zhark
-
AFNI Message Board
I have written a simple program 3dPval which will read a dataset, convert the volumes marked as statistics to p-values (leaving alone the non-statistic volumes), and write the result out.
This program will in the next build of AFNI, which might happen as early as tonight.
by
Emperor Zhark
-
AFNI Message Board
You can use an all-zero column in 3dREMLfit if you also use the option '-GOFORIT'
However, a simpler way might be to use the "semi-hidden" option '-matim' (which you can read about at this link):
Quote-matim M = Read a standard .1D file as the matrix.
** N.B.: You can only use 'Col' as a name in GLTs
with these nonsta
by
Emperor Zhark
-
AFNI Message Board
It is something I forgot about, and should do (I noticed this gap the other day myself). However, I've got so much else to think about for the next 3 months that I just don't know when I'll get around to it.
by
Emperor Zhark
-
AFNI Message Board
At present, no. Making Clusterize drive-able is on my list of things to do.
by
Emperor Zhark
-
AFNI Message Board
From this day forward, attempting to plot -xjpeg with a matrix larger than 3333 (in either direction) will result in a warning message and no JPEG output.
by
Emperor Zhark
-
AFNI Message Board
Claudio,
It's not clear from your first question whether you are correlating betas across subjects between the two tasks, or are correlating results in 1 subject across different events in the 2 tasks.
In either case, you have to create 2 datasets (via 3dbucket or 3dTcat), that contain the beta "time series" (or "subject series") -- one dataset for each task. Then
by
Emperor Zhark
-
AFNI Message Board
If you type the command
3dinfo -verb stats.Imagination001_REML+orig
you will get a screen printout of the information in the dataset header. The information for each volume (=sub-brick) in the dataset will include the type of statistic it is (t or F), and the degrees of freedom.
With that information, you can convert the voxel values from statistics to nominal p-values via 3dcalc. For e
by
Emperor Zhark
-
AFNI Message Board
3dttest++ now can analyze a single dataset and compare it to a collection of reference datasets. The result is a t-statistic in each voxel that measures the difference between the singleton dataset and the mean of the reference datasets, normalized by the standard deviation of the reference datasets and by a scale factor that makes the null-hypothesis distribution be the Student t. In addition,
by
Emperor Zhark
-
AFNI Message Board
The problem seems to be introduced by the compiler optimization. When I add a debugging printout statement to print the option out before parsing it, the problem goes away -- even when the debug option is not invoked, so the printout never occurs!
I've seen stuff like this a few times before, but it is always peculiar. Anyway, I will start a rebuild and new binaries will be available by
by
Emperor Zhark
-
AFNI Message Board