I was wondering if any more progress had been made to implementing the “residuals” option in 3dLMEr? This was brought up in the following thread a few months ago (https://afni.nimh.nih.gov/afni/community/board/read.php?1,162874,163747#msg-163747), but I know how things get busy! - and I see the option is still commented out in github. I updated R to 4.0 and now 3dLME will not run, and I need a way to get residuals so that I can output into 3dFWHMx -ACF to correct for multiple comparisons. If this option is not available in 3dLMEr, any other suggested strategies for determining cluster size thresholds? Thank you!
Option -resid in 3dLMEr has been turned on in the updated version on github, and will be available in the next AFNI release.
That’s great! And just to confirm, 3dLMEr runs well with R 4.0 versions, correct? (I know there were problems with 3dLME.) Also, do you have an estimate of when this will be available, and if there will be a notice sent out when this is released? Thanks so much!
just to confirm, 3dLMEr runs well with R 4.0 versions, correct?
There might be some issue with R 4.0 on Mac, but it should be fine on Linux machines.
do you have an estimate of when this will be available, and if there will be a notice sent out when this is released?
If you cannot wait for the next official release, directly download the following and put it in your AFNI binary directory: https://afni.nimh.nih.gov/sscc/staff/gangc/pub/3dLMEr.R
Thank you very much for your help!
Just a follow up on this thread,
I’m using 3dLMEr Version 0.1.0 (April 11, 2021) and there is now the -resid option but I keep getting this error when I use it :
Package snow loaded successfully! Error in Stat[, , kk, 1] <- parApply(cl, inData[, , kk, ], c(1, 2), runLME, : number of items to replace is not a multiple of replacement length Execution halted
The model work well when I don’t use it though. (I am running on CentOS 7.9, with AFNI_21.1.08 ‘Domitian’ using R version 3.6.0).
Any idea what can cause this ?
Thanks in advance,