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
You can actually do this now.
Step 1: create a directory, say $HOME/aglobal
Step 2: copy any file(s) you want to be always present into that directory, for example, MNI152_2009_template.nii.gz
Step 3: in your $HOME/.afnirc file, add the line (to the ***ENVIRONMENT section)
AFNI_GLOBAL_SESSION = /Users/rwcox/aglobal
When you start AFNI, the files in this global session will always appea
by
Bob Cox
-
AFNI Message Board
The new ETAC method for cluster-thresholding is now ready for use, in program 3dttest++.
The slides from my presentation today (26 July 2017) at the NIH are available: https://drive.google.com/file/d/0B6Sn11ZTBrhvOGJLek9iWWRkOTg/view?usp=sharing
The help for 3dttest++ contains more details about how to use this method: https://afni.nimh.nih.gov/pub/dist/doc/program_help/3dttest++.html
ET
by
Bob Cox
-
AFNI Message Board
The thresholding process envisioned and implemented in 3dClustSim has 2 steps:
(1) Per-voxel thresholding at some p-value. Smaller p-values are more strict: fewer voxels will "survive".
(2) From the "surviving" voxels, form spatially contiguous clusters. Delete those that are smaller than the cluster-size threshold.
Consider the case where you lower step 1's per
by
Bob Cox
-
AFNI Message Board
Oops. I've now fixed this (one extra line of code for the -singletonA case). Source code is checked in, but won't be compiled into the binaries until later this week.
by
Bob Cox
-
AFNI Message Board
3dFWHMx no longer shows the "classic" (Forman 1995 -http://onlinelibrary.wiley.com/doi/10.1002/mrm.1910330508/full) estimate of the Gaussian FWHM of the noise smoothness. This change means that anyone using the Gaussian-shape model for the spatial correlation of the FMRI noise will have trouble doing so in AFNI -- the purpose of this change is exactly to cause such trouble.
I have t
by
Bob Cox
-
AFNI Message Board
Rick is correct -- it would be tricky in the code to put a -stim_times regressor in the baseline model, so I didn't make that possible. As it really only affects the Full-F statistic (again, Rick is correct -- imagine that!), I decided it wasn't worth the programming effort.
by
Bob Cox
-
AFNI Message Board
It is very hard to say, with the limited error message output by auto_warp.py. We can't test this run here, since we don't have a similar system. One possibility is that the program ran out of memory and was killed by Linux (which is nearly silent when it does something so terrible). This situation might be possible if the two different computers have different amounts of RAM. It di
by
Bob Cox
-
AFNI Message Board
First, the '-volreg' option can only be used when you input exactly the '[1..6]' columns (roll pitch yaw dS dL dP). If you input anything else, the labels will be wrong (as you've seen). You can only get what you want by putting the desired labels in manually, as in
1dplot ... -ynames dS dL dP - Task3dmotion.1D'[4..6]'
where the standalone '-' m
by
Bob Cox
-
AFNI Message Board
For almost any purpose, Pearson correlation is what you want. The others are experimental.
Spearman would be used if you expect a lot of outliers. Usually it will be almost identical to Pearson.
Taub's k is really useful only when the data consists of a few levels (such a numbers 1, 2, 3, 4, or 5), not a continuously variable value.
Quadrant correlation is never useful, and is the
by
Bob Cox
-
AFNI Message Board
An uncompressed dataset in .HEAD/.BRIK format is mapped into memory from disk, which can work with very large files. Otherwise, you must have more memory (RAM) available than the size of the file. Memory mapped large files might work, but they will tend to be very slow if there isn't enough RAM to page them into memory.
So really, the answer to your question depends on the amount of mem
by
Bob Cox
-
AFNI Message Board
You ask difficult questions, Paul. I'm planning on addressing this to some extent in the future, although my enthusiasm for clustering is about equal to 1/(first Skewe's number).
by
Bob Cox
-
AFNI Message Board
I looked at these messages again last night. There are 3 issues here:
Your version of AFNI needs updating. ETAC was updated in late May 2017. For some reason, all the .minmax.1D files didn't get read. That is the substance of the ERROR messages. These are minor errors, and the program continues without further comment. If this problem re-occurs with an updated version of AFNI, I'd
by
Bob Cox
-
AFNI Message Board
If you want to transform a T1-weighted dataset to MNI space and at the same time skull strip it, the @SSwarper script can do the job for you. It applies a combination of 3dSkullStrip and 3dQwarp to do these calculations.
by
Bob Cox
-
AFNI Message Board
It forms the sin() and cos() regressors in the rejected bands, and those are included in the matrix whose columns are to be projected out of the dataset. This method isn't a "filter" in the sense of "Butterworth" or other time-domain FIR or IIR types of objects with which you might be familiar.
by
Bob Cox
-
AFNI Message Board
You have a peculiar mix of dates on the programs you are running. 3dttest++ is Apr 12 2017, but 3dClustSim is Mar 21 2016. These programs are designed and modified to work together, so if one gets updated but the other does not, then bad things will happen.
You need to properly upgrade your AFNI binaries, so they are all synchronized.
by
Bob Cox
-
AFNI Message Board
I can only think you have an out-dated version of AFNI. What the output of the command afni --ver ? If the version number is below AFNI_17.1.00, you need to update your AFNI binaries.
by
Bob Cox
-
AFNI Message Board
Per-voxel p-values are computed from the statistical values stored in a dataset (for example, the t-statistics).
If you want to convert a dataset statistical volume to p-values, the program 3dPval can do this for you.
by
Bob Cox
-
AFNI Message Board
P.S. -- in my browser, his image is stretched vertically. That's not the way he really is.
by
Bob Cox
-
AFNI Message Board
Justin Rajendra has joined the AFNI gang, and is already making good things -- for example, an updated version of uber_subject.py, and making AFNI work with Docker. You will likely see him responding to messages here, and don't be worried -- he speaks ex cathedra like the rest of us .
I've (hopefully) attached a picture of Justin. Don't be alarmed -- despite appearances, he
by
Bob Cox
-
AFNI Message Board
If I were you, I'd look at the xyz points in the original (warped) space where they were created, then look at the transformed points in the unwarped space, and see if they are in the same anatomical location -- using the warped anatomical as the underlay in the first case, and the unwarped anatomical as the underlay in the second case. This procedure will quickly inform you if the transfor
by
Bob Cox
-
AFNI Message Board
Program 3dNwarpXYZ can transform coordinate triples (x,y,z) given a nonlinear warp. Please read the help carefully to make sure you are using this program correctly, because it is easy to specify the nonlinear warp incorrectly.
by
Bob Cox
-
AFNI Message Board
If you are aligning to the template MNI152_2009_template.nii.gz (part of AFNI binaries), then you can use the #4 sub-brick of that as a gray matter plus CSF mask. You can align to that template using the @SSwarper script supplied with AFNI, for example.
In general, a gray matter mask can only be useful if reasonably high-order nonlinear registration (e.g., 3dQwarp) is used to align the brain
by
Bob Cox
-
AFNI Message Board
I had assumed you were more familiar with 3dcalc.
Your mistake is twofold. First, you used '-a' to define the input, whereas in my formula, I used the variable name 'x', so you should use '-x' to give input dataset name (or change the 'x' in the stat2cdf function to 'a'). Second, you multiplied the function calls by 'a', which is no
by
Bob Cox
-
AFNI Message Board
Yes, although the 3dcalc expression is a little clumsy. The method is to (1) convert the z-statistic to its CDF value and then (2) convert that CDF value to a t-statistic with the desired number of degrees of freedom (DOF) -- you need to supply the DOF parameter.
The expression is like this
"cdf2stat(stat2cdf(x,5,0,0,0),3,DOF,0,0)"
where DOF is the degrees of freedom paramete
by
Bob Cox
-
AFNI Message Board
Email me the matrix file and the exact 3dDeconvolve command you used to create the matrix. I'll take a look at that.
robertcox *at* mail.nih.gov
by
Bob Cox
-
AFNI Message Board
Why do you say "The differences are minor, however they pose some concerns for subsequent analyses"?
Minor differences should not produce major changes in the results of subsequent analyses. If minor differences at one stage make a big difference in the final brain mapping results, then there is something else wrong with the analysis -- such as bad data or other bad inputs (e.g., a
by
Bob Cox
-
AFNI Message Board
The simplest possible script is to create a file with 1 line per command you want to execute.
3dROIstats -1DRformat -quiet -mask directory1/inputfile1 > outputfile1.1D
3dROIstats -1DRformat -quiet -mask directory2/inputfile2 > outputfile2.1D
and so on -- create the first line, copy it, and then edit it into the format you need. If you are good with an editor and if your filenames ha
by
Bob Cox
-
AFNI Message Board
What you need to do is called "shell scripting" or "shell programming". An introduction for scripting with the C shell (the one we AFNI-oids prefer) is here:
http://www-cs.canisius.edu/ONLINESTUFF/UNIX/shellprogramming.html
For your particular problem as posed, one solution involves creating and using shell variables, in a loop.
For example
set flist = ( `cat s
by
Bob Cox
-
AFNI Message Board
There is a slight error in 3dTproject -- not in the algorithm (as far as I know) but in the error message you got.
It says "235 retained time points" -- that is, the number left after censoring -- but 235 is the TOTAL number of time points, and 166 is the number left after censoring. But the comparison that produces the error is (correctly) between the number of regressors being rem
by
Bob Cox
-
AFNI Message Board