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 181 - 210 of 2632
Chuck,
Are Incentive, Magnitude, and Outcome 3 within-subject factors each of which has 2 levels? One possibility is that you didn't provide enough number of subjects. In fact, there are only 2 subjects in the Control group.
by
gang
-
AFNI Message Board
> S#0 will give me the intercept (and is associated with the major component of the hemodynamic response,
> as the first component from SPMG2, right? ) and S#1 will give me the slope, or the signal modulated by the
> parametric modulation of the function component (please correct me if I am wrong).
Yes, that's right.
> I still don't know what will represent the outpu
by
gang
-
AFNI Message Board
Julie,
> We ran 3dMVM with R version 3.2.4 and AFNI version AFNI_16.0.16. It does NOT run with R v. 4.1.1 (or 3.6.3) and AFNI v. AFNI_21.2.04.
Do you mean that the same 3dMVM script worked fine with the former but not with the latter?
What is the output of the following command when you run it on the latter system?
afni_system_check.py -check_all
by
gang
-
AFNI Message Board
> When the SPMG1 function is used, ... I get the sub-briks under this names: S#0 and S#1. Which beta contain the effect
> of parametric modulation and which doesn't?
S#0 is the intercept while S#1 is the slope (modulation).
> When the SPMG2 function is used, ... I get the sub-briks under this names: S#0, S#1, S#2 and S#3. Which beta contain
> the effect of parametric mod
by
gang
-
AFNI Message Board
Philipp, I don't see any strong rationale for the averaging process. So I would consider dealing with all runs together (your option 3).
by
gang
-
AFNI Message Board
Yiyang,
Since your data are in NIfTI format, almost all AFNI programs can directly work on your data (2D or 3D).
by
gang
-
AFNI Message Board
Hi JW,
It would be interesting to try out MBA and see how it performs.
> On the MBA page, all the examples have two ROIs.
Practically the modeling approach with MBA can take a few up to 100 regions. Those two columns in the help are meant to indicate the two regions for each region pair (matrix element), not the number of regions allowed.
> have a full and symmetric matrix (z sc
by
gang
-
AFNI Message Board
> I am interested in performing within and between network analysis so I assume I need to use MBA?
Could you describe the data structure? How many ROIs do you have? For each subject, do you have a full (and symmetric) matrix for those ROIs or do you have missing data for the matrix? What do you mean by "within and between network analysis"?
by
gang
-
AFNI Message Board
> Before 3dLSS existed the only way I knew to do that was to run 3dDeconvolve once for each stimulus, modeling all other stimuli separately.
Unless your trials are really too close to each other, why not use option -stim_times_IM in 3dDeconvolve and model the trials simultaneously (instead of separately)?
by
gang
-
AFNI Message Board
Leah,
> Are the AFNI ACF blur estimates supposed to be using stats.s####_REML file as the final dataset?
Conceptually ACF/smoothness estimates should reflect the spatial relatedness in pure "noise" because you're supposed to control "false positives" assuming there is no signal in the data under the traditional statistical framework. So, it may make more sense to
by
gang
-
AFNI Message Board
> Would it be reasonable to make a separate directory called residuals and run 3dmvm twice with the table split in half, to generate
> residuals form the top and bottom half of the table and then average the two residuals volumes and feed it into 3dFWHMx?
If you cannot find a computer with a larger memory size, you can try what you're suggesting. It is not ideal, but the approximat
by
gang
-
AFNI Message Board
> the resulting individual average beta values from 3dmaskave plotted with the behavioral scores don't match the polarity of the clusters from the regression.
In what sense they don't match? Could you show (1) the scatter plot of contrast vs behavioral score at one ROI, and (2) the regression coefficient from 3dttest++ at that ROI?
by
gang
-
AFNI Message Board
Hi Will,
> In trying to prep data for an RSA analysis, your 3dLSS program is wonderfully efficient.
Efficient in what sense? Compared to 3dDeconvolve/3dREMLfit?
> 3dLSS does not have an option for outputting t-maps.
From the modeling perspective, 3dLSS is inferior to 3dDeconvolve/3dREMLfit. Essentially 3dLSS is a band-aid approach to fixing the potential numerical problem of hig
by
gang
-
AFNI Message Board
> my script was: a1b1-a1b2-a2b1+a2b2
If I didn't misread your original specification, it was actually
a1b1-a1b2+a2b1-a2b2
> Your suggestion was: a1b1-a1b2-a2b1-a2b2
I meant to show
a1b1-a1b2-a2b1+a2b2
but I forgot to change some of the signs after copy and paste.
by
gang
-
AFNI Message Board
Philipp, that web page is out of date. it's better to follow examples 5, 9, and 10 in the help of afni_proc.py
by
gang
-
AFNI Message Board
> Does our setup look correct for a 3way design, 2 within and 1 between subjects factor?
It looks fine. You could set a finer model
-model "pain*faces*group" \
> Using 3dLME, does one just enter the available datasets and 3dLME recognizes what data points are missing? Or do you we have to code it a certain way (adding NA or ".")
Just provide the complete case
by
gang
-
AFNI Message Board
> To my best understanding, fat_mvm_prep.py and fat_mvm_scripter.py should be used to perform group analysis.
Yes, I believe that you can use those programs. Another alternative is RBA (https://afni.nimh.nih.gov/pub/dist/doc/program_help/RBA.html).
by
gang
-
AFNI Message Board
Lingyan,
> I have two factors, a and b, each has two levels (a1,a2 and b1,b2). In total I have four conditions (a1b1,a1b2,a2b1,a2b2).
> In each of these conditions, I test 9 different touch frequencies stimulation
Actually your experiment seems to be a 2 x 2 x 9 design. The specifications for the two main effects look fine, but the one for interaction
+a1b1_f1 +a1b1_f2 ... +a1b1_f
by
gang
-
AFNI Message Board
Those warnings about FDR simply indicate that the associated statistical evidence is weak in terms of the false discovery rate. In other words, that is just how much information your data could provide based on the current model.
You can also ignore the following message (a numerical indicator for some statistical construct at some voxel(s)).
> Warning message:
> In summary.Anova.ml
by
gang
-
AFNI Message Board
> I first centered around the mean of each session. Then in the 3dLME script I am testing what the effects would be
> if Movement was not happening, i.e. =0 (which I guess would be the default anyway). Is this interpretation correct?
Yes, your interpretation seems correct. If you meant the default of 3dLME for centering, the center value is actually the overall average, not 0.
> I
by
gang
-
AFNI Message Board
OK, I suspect that your AFNI version is a little old. This was a problem at some point, but it got fixed later. On the other hand, practically you can just ignore or delete those extra 1,800 sub-bricks in results+tlrc
by
gang
-
AFNI Message Board
> when I run the above the resulting "results_no_smooth+tlrc." file(s) have 1804 sub-briks. I.e. all the subjects plus the stats.
Do you mean that you got two separate output files: one with those statistical results while the other contains statistical results plus the residuals? If so, that is strange. I never tested the case with a 4D file as an input. So, if the problem persis
by
gang
-
AFNI Message Board
I jumped through many hoops, but I don't recall installing the few dependencies you mentioned. Instead, I executed the following to resolve the issue related to the R packages such as "units" and "sf" in your error list:
sudo add-apt-repository ppa:ubuntugis/ubuntugis-unstable
sudo apt-get update
sudo apt-get install libudunits2-dev libgdal-dev libgeos-dev libproj-
by
gang
-
AFNI Message Board
I just had a similar headache because of some system updates on Ubuntu 20.04. After a long meandering path, I eventually managed to sort out all the hiccups.
Try the following command inside R:
install.packages("Rcpp", dependencies = TRUE, INSTALL_opts = '--no-lock')
install.packages("brms", dependencies = TRUE, INSTALL_opts = '--no-lock')
Paste
by
gang
-
AFNI Message Board
> 1) Should I loop through this sub-brik with 3dFWHMx, where the loop index is used to give 3dFWHMx one sub-brick at a time or can I just give it the full 4D-volume?
Do it on the 4D data (residuals for the model).
> 2) Should I use any of these option?
Those options do not seem to be applicable in this case.
> 3) Should I for structural data still use -ACF or is classical -fw
by
gang
-
AFNI Message Board
Robin,
A couple of suggestions. Use the option -resid in 3dMVM to obtain the residuals. Then estimate the spatial smoothness with 3dFWHMx, and proceed with usual ritual. On the other hand, the cluster approach is a Band-Aid solution anyway, so here is my radical opinion:
by
gang
-
AFNI Message Board
Could you provide some context? How is the experiment structured and designed? What kind of model (e.g., regressors) are you implementing through 3dDeconvolve?
by
gang
-
AFNI Message Board
I don't have an answer for the 3dLSS question, but my suggestion is that, unless there is a strong multicollinearity problem, using 3dDeconvolve is much more preferable to dealing with 3dLSS.
by
gang
-
AFNI Message Board