Hi there,
I'm attempting to use 3dLME to model the interaction between two within-subject categorical variables and the linear and quadratic effect of age (this is a longitudinal study). When running the following model without specifying contrasts, it is estimated without problem.
3dLME -prefix SelfOverTime_3Ts -jobs 8 \
-model "target*domain*age_c+target*domain*age_c2" \
-ranEff '~1' \
-SS_type 3 \
-qVars "age_c,age_c2" \
-qVarCenters "0,0" \
-mask /path/mask.nii \
-dataTable \
Subj target domain age_c age_c2 InputFile \
s005 self academic -3.09016393442623 9.54911314162859 /path/s005_t1_con_0001.nii \
s005 self social -3.09016393442623 9.54911314162859 /path/s005_t1_con_0002.nii \
s005 other academic -3.09016393442623 9.54911314162859 /path/s005_t1_con_0003.nii \
s005 other social -3.09016393442623 9.54911314162859 /path/s005_t1_con_0004.nii \
s005 self academic -0.0273972602740002 0.000750609870521312 /path/s005_t2_con_0001.nii \
s005 self social -0.0273972602740002 0.000750609870521312 /path/s005_t2_con_0002.nii \
s005 other academic -0.0273972602740002 0.000750609870521312 /path/s005_t2_con_0003.nii \
s005 other social -0.0273972602740002 0.000750609870521312 /path/s005_t2_con_0004.nii \
s005 self academic 3.6602739726027 13.3976055545127 /path/s005_t3_con_0001.nii \
s005 self social 3.6602739726027 13.3976055545127 /path/s005_t3_con_0002.nii \
s005 other academic 3.6602739726027 13.3976055545127 /path/s005_t3_con_0003.nii \
s005 other social 3.6602739726027 13.3976055545127 /path/s005_t3_con_0004.nii \
However, when adding contrasts to the model, it fails to run (standard error message). All of the necessary packages were installed, and I believe the contrasts are correctly specified. I've tried to run a model with the following contrasts:
-num_glt 6 \
-gltLabel 1 'self-other' -gltCode 1 'target : 1*self -1*other' \
-gltLabel 2 'social-academic' -gltCode 2 'domain : 1*social -1*academic' \
-gltLabel 3 'age self-other' -gltCode 3 'target : 1*self -1*other age_c : ' \
-gltLabel 4 'age^2 self-other' -gltCode 4 'target : 1*self -1*other age_c2 : ' \
-gltLabel 5 'age social-academic' -gltCode 5 'domain : 1*social -1*academic age_c : ' \
-gltLabel 6 'age^2 social-academic' -gltCode 6 'domain : 1*social -1*academic age_c2 : ' \
Since that model failed though, I also tried running a model with this simple contrast alone:
-num_glt 1 \
-gltLabel 1 'self' -gltCode 1 'target : 1*self' \
Could you provide any guidance as to what I might be doing wrong?
Thank you very much for your help!
