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
Page 6 of 6
Pages: 23456
Results 151 - 180 of 180
I thought I was clear -- if you do not use the covariate, the labels or lack thereof do not affect the numerical results. (It would be very hard to imagine how they COULD affect the results.)
by
Emperor Zhark
-
AFNI Message Board
The warning is there because the dataset names/labels will also be used when looking up covariates; that is, if you used the '-covariates' option. If that were the case, the '-covariates' option would be very unhappy with you (since it would be unable to tell which line in the covariates table went with which input dataset). Otherwise, what you got is just a warning message
by
Emperor Zhark
-
AFNI Message Board
The afni server was down this Easter weekend, due to disk drive problems. Sorry about this problem, and hope it didn't ruin your celebrations.
by
Emperor Zhark
-
AFNI Message Board
My suggestion was to change the penalty factor (last number on the given command line) from '0' to '-1'. This gave decent results for me. The automatic penalty factor finder sometimes runs amok, and produces too-smooth results.
by
Emperor Zhark
-
AFNI Message Board
It's feasible -- e.g., to map one set of widgets while unmapping another, so only 1 is visible at a time -- something like this is done for the widgets below the 'Clusterize' menu. But you would only see one of the 'Define Overlay' panels at a time, which could be clumsy. But screen space would be saved -- which is important for a laptop, say.
by
Emperor Zhark
-
AFNI Message Board
How would you imagine the "two overlays" thing would work? At present, the widgets to control how the overlay dataset is converted to colors (threshold slider, sub-brick index choosers, etc.) take up a lot of space. Doubling these would be ugly and hard to use. This conceptual issue is one thing that's held me back from going down this rabbit hole.
by
Emperor Zhark
-
AFNI Message Board
Email me (robertcox *at* mail.nih.gov) your seed_TS.1D and model_TENT_HRF.1D files and I'll take a look at the issue. More than that I can't say now -- I wrote 3dTfitter years ago, and haven't actually used it for a while now.
by
Emperor Zhark
-
AFNI Message Board
Just for you Paul ---
New environment variable AFNI_SLAVE_THROLAY will initialize the slaving relationship between the overlay sub-brick index and the threshold index (which can later be changed in the GUI from the Index popup choose or the threshold slider popup menu). You can
set this variable (before starting AFNI, in any of the usual ways) to
'==' *or* 'OLay'
by
Emperor Zhark
-
AFNI Message Board
Well, I was reminded of this today (by someone who shall remain un-named), and so I'm starting to look into it again. I managed to find my notes on the subject, so that's a start. Next up I'm getting some data from someone here at the NIH to try out the field map approach.
by
Emperor Zhark
-
AFNI Message Board
Answer #1 == Right click on the 'Index' label (at the left middle of the main AFNI controller) to get a chooser that lets you control these things.
Answer #2 == Right click on the label at the top of the threshold slider for yet another such menu.
by
Emperor Zhark
-
AFNI Message Board
In the AFNI user interface, there is only a clumsy way to do this, via the InstaCalc controls.
Open the Define Overlay control panel In the upper right corner is a drop-down menu with the word "Clusters" -- click on that and switch it to "InstaCalc" Below will appear a button "Setup ICalc" -- press that button A biggish control panel will open up to let you
by
Emperor Zhark
-
AFNI Message Board
So fdrval has been modified such that if a too-small q value is input via the '-qinput' option, then the resulting threshold will be as large as or larger than the largest threshold value in the specified dataset volume.
by
Emperor Zhark
-
AFNI Message Board
Well, I could have it return an error if q gets too small, or I could have it return a threshold value so large that no voxels survive. Which is better?
by
Emperor Zhark
-
AFNI Message Board
The next bootcamp at the NIH will be in October 2014, but I don't have the dates handy at this moment. There will also be some workshops at U Miami, Yale, and Dartmouth.
by
Emperor Zhark
-
AFNI Message Board
The message Killed when a program suddenly stops means that the operating system shut the program down -- usually because the program asked for too much memory and the system needed to get memory back to function properly.
There is nothing the program can do to stop this (except use less memory).
by
Emperor Zhark
-
AFNI Message Board
When I added the 'Set q-value' chooser, I altered the left-click => 'Set p-value' to be the scroll wheel pops up the p-value (scrolling one way) or the q-value (scrolling the other way) chooser. Also, this works both on the top AND bottom labels in the threshold slider column.
by
Emperor Zhark
-
AFNI Message Board
It is not clear to me what this sentence means:
The averaged results are not what I expected and they look similar to me.
I take the last clause means that the averaged results from fnirt and 3dQwarp are similar. But does the first clause mean that you don't expect them to be similar, or that the results are not what you expect for some other reason?
Assuming that it is the similar
by
Emperor Zhark
-
AFNI Message Board
In recent days, I've added a couple of features to give the AFNI user some extra control over time stepping. In particular, the 'Index' label in the leftmost column of the main AFNI control window now has a right-click popup menu attached -- I've tried to attach a JPEG of it. There are 3 controls in this popup:
Index Step :: the amount the time index steps forward/backwa
by
Emperor Zhark
-
AFNI Message Board
I believe that mcflirt returns rotation angle estimates in radians, while AFNI uses degrees. Thus the AFNI angles numbers will be considerably larger than the mcflirt angle numbers (57.295780 times larger, approximately).
by
Emperor Zhark
-
AFNI Message Board
3dUniformize is kind of old, and was written by someone no longer involved in AFNI development. A newer program is 3dUnifize, which may serve your purposes better. Have you tried it?
by
Emperor Zhark
-
AFNI Message Board
OK, after working a little with Rick on his Linux system, the issue comes down to the way that AFNI detects if the 'Alt' (Linux) or 'Command' (Mac) key is pressed down while the scroll wheel is scrolling. This detection is done by examining the 'modifier state mask' during the scroll 'event'. The mask is usually set to Mod1 for the Alt key on Linux and to
by
Emperor Zhark
-
AFNI Message Board
Interesting. By ssh-ing into a Linux machine, I see some of this behavior but not all -- when displaying on my Mac. So it must be a combination of the X-server (on the Mac) and the X11/Motif libraries (on the Linux box) giving the stuff I see -- in particular, I do get scrolling with the scroll wheel over the image. I also see, when holding down the Alt/Command key, the Overlay threshold slide
by
Emperor Zhark
-
AFNI Message Board
Your AFNI's behavior is peculiar. In my version (on a Mac), the scroll wheel moves through slices/images easily when the mouse cursor is over the image, but does nothing when it is over the slider below the image. The source code only looks for scroll wheel events in the image, image intensity bar, and graph windows.
Without a system like yours to play with, it's kind of hard to fi
by
Emperor Zhark
-
AFNI Message Board
I think you'll have to use 2dImReg for a single slice dataset.
by
Emperor Zhark
-
AFNI Message Board
OK, so I've patched the code so if the WEIRDNAME.nii file doesn't exist, it will retry with WEIRDNAME.nii.gz -- that should soothe your problem.
The source code is available now. I will rebuild binaries tonight -- which one do you want? I can rebuild that one right away, if practicable.
by
Emperor Zhark
-
AFNI Message Board
The "weird" names are the unique name strings generated to prevent conflicts if you have multiple copies of the same program running in the same directory at once.
However, in the usual case (where the 3dallineate-d file is to be kept), the "weird" output filename is renamed (inside 3dQwarp) to PREFIX_allin.nii
Now that I write the above sentence, I see the problem. 3dA
by
Emperor Zhark
-
AFNI Message Board
Have you tried running with the '-verb' option? It will print out more information as the programs proceed, and perhaps give a clue.
I can't duplicate this problem here, trying to keep the same set of options you have as much as possible.
by
Emperor Zhark
-
AFNI Message Board
Please note that as of NOW NOW NOW we will no longer be building AFNI binaries for Mac OS X 10.5. My desktop (the build machine for 10.5) was finally upgraded to 10.8, and so from now on I will be building 10.8 binaries instead.
There will be 2 packages -- macosx_10.8_gcc and macosx_10.8_icc
The _gcc package is available now (I hope it works for you), but the _icc package is not -- I'
by
Emperor Zhark
-
AFNI Message Board
Nobody knows the answer to that question -- we can just guess.
If you have physiological data (pulse ox, etc), then you would help things by removing the RetroTS covariates -- as is possible in the afni_proc.py pipeline.
If you have 2 groups of subjects (patients and controls say), who are exclusively at the different TRs (say all patients at TR=2, all controls at TR=1.5), then you will be
by
Emperor Zhark
-
AFNI Message Board
It should be there now.
Shortly, I'll be upgrading that machine to OS X 10.8, and hopefully will be providing 10.8 icc builds from then on (and 10.5 support will be ended).
by
Emperor Zhark
-
AFNI Message Board
Page 6 of 6
Pages: 23456