3dMVM glt error: "Inappropriate coding in test"

AFNI Message Board
1

Hi AFNI experts,

I’m hoping to use 3dMVM to conduct a 2x2x2 ANOVA with unequal subjects across two between-subjects conditions. The between subjects factor is Suicide Attempt History: Yes vs No. The within subjects factors are two experimental manipulations: emotional valence of experimental stimuli (Negative v. Neutral) and Attention Shifting (trials requiring attention shift v. repeat trials focusing on the same target). Using the code and data table format below, I’m getting an error message: "Inappropriate coding in test No.1! " In GLT 1, I’m hoping to find the difference between activation for negative (v. neutral) images specific to repeat trials for suicide attempters only. Any leads as to what I need to change would be greatly appreciated! Many thanks in advance.

3dMVM -prefix 3dMVMtry -jobs 1
-bsVars SAs
-wsVars "valshr"
-mask mask_overlap.7+tlrc
-qVars “age”
-num_glt 3
-gltLabel 1 SAtt_negRneuR -gltCode 1 'SAs : 1SAtt shr : 1repeat val : 1neg-1neu’
-gltLabel 2 NAtt_negRneuR -gltCode 2 'SAs : 1NAtt shr : 1repeat val : 1neg-1neu’
-gltLabel 3 intSA_negRneuR -gltCode 3 'SAs : 1SAtt - 1NAtt shr : 1repeat val : 1neg-1neu’
-dataTable @/Users/imaging1/Desktop/3dMVMtry4.txt \

Excerpt from the Data table:
(two notes: 1) I can’t make the tabs display properly in this copy, but they do appear ok in the actual text file; 2) I deliberately posted age here as a 3-digit number to deidentify the data)

subj SAs val shr age confl InputFile
3 NAtt neg shift 216 co stats.subject3_REML+tlrc[1]
3 NAtt neg shift 216 noco stats.subject3_REML+tlrc[6]
3 NAtt neg repeat 216 co stats.subject3_REML+tlrc[11]
3 NAtt neg repeat 216 noco stats.subject3_REML+tlrc[16]
3 NAtt neu shift 216 co stats.subject3_REML+tlrc[21]
3 NAtt neu shift 216 noco stats.subject3_REML+tlrc[26]
3 NAtt neu repeat 216 co stats.subject3_REML+tlrc[31]
3 NAtt neu repeat 216 noco stats.subject3_REML+tlrc[36]
7 NAtt neg shift 249 co stats.subject7_REML+tlrc[1]
7 NAtt neg shift 249 noco stats.subject7_REML+tlrc[6]
7 NAtt neg repeat 249 co stats.subject7_REML+tlrc[11]
7 NAtt neg repeat 249 noco stats.subject7_REML+tlrc[16]
7 NAtt neu shift 249 co stats.subject7_REML+tlrc[21]
7 NAtt neu shift 249 noco stats.subject7_REML+tlrc[26]
7 NAtt neu repeat 249 co stats.subject7_REML+tlrc[31]
7 NAtt neu repeat 249 noco stats.subject7_REML+tlrc[36]
…
…
…

2

An empty space is needed to separate each weight from the previous effect (and don’t leave a space between the negative sign and the value). That is, the following three lines

-gltLabel 1 SAtt_negRneuR -gltCode 1 ‘SAs : 1SAtt shr : 1repeat val : 1neg-1neu’
-gltLabel 2 NAtt_negRneuR -gltCode 2 ‘SAs : 1NAtt shr : 1repeat val : 1neg-1neu’
-gltLabel 3 intSA_negRneuR -gltCode 3 ‘SAs : 1SAtt - 1NAtt shr : 1repeat val : 1neg-1*neu’ \

should be

-gltLabel 1 SAtt_negRneuR -gltCode 1 ‘SAs : 1SAtt shr : 1repeat val : 1neg -1neu’
-gltLabel 2 NAtt_negRneuR -gltCode 2 ‘SAs : 1NAtt shr : 1repeat val : 1neg -1neu’
-gltLabel 3 intSA_negRneuR -gltCode 3 ‘SAs : 1SAtt -1NAtt shr : 1repeat val : 1neg -1*neu’ \

3

Hi Gang,
Thank you so much for looking over this so quickly! I made the change and for some reason, I keep getting the same error. I checked for typos a couple of times, and also tried the code with a single GLT (in which case, I get a different error: “Error in outList[[1]][[n]] : subscript out of bounds | Calls: process.MVM.opts → gl_Constr”). I tried putting the sub-brick indicator in the data table in quotation marks ‘[1]’ instead of [1], which still didn’t work. Let me know if there is something else to try - and thank you so much again for your help!
Irena

4

Hi,

It seems like we have very similar problems. It looks like 3dMVM will run without the GLTs fine, so it seems like some thing is not right with my, and possibly your glt coding.