GLMFlex to GLMFlexFast

[Haris Styliadis]

Hello Aaron, you recently suggested updating my script to use GLMFlexFast instead of GLMFlex.

I have read several discussion and followed every example I came across.

I would really appreciate it if you had a look at what I have prepared. I am attaching both the new code. The aim here is to model a 2(Group1 and Group2)x2(high valence and low valence)x2(high arousal and low arousal) anova and estimate the following effects and interactions.

1. Group by Valence by Arousal -------------------------F-test 
2. Valence by Arousal Group1----------------------------F-test for Group1
3. Valence by Arousal Group2----------------------------F-test for Group2
4. Valence by Arousal---------------------------------------F-test
5. Group by Arousal ----------------------------------------F-test
6. Arousal Group1-------------------------------------------t-test for Group1
7. Arousal Group2-------------------------------------------t-test for Group2
8. Arousal Effect---------------------------------------------t-test 
9. Group by Valence ---------------------------------------F-test
10. Valence Group1------------------------------------------t-test for Group1
11. Valence Group2------------------------------------------t-test for Group2
12. 'Valence Effect -------------------------------------------t-test 

I am a bit confident for the valence and arousal main effects (F and T-tests) but need help for the interactions. 

I have run only the comparisons for the valence and arousal main effects but I get the following error message

Index exceeds matrix dimensions. at line 19

          

             

              

[Haris Styliadis]

Thanks for the support!

[Aaron Schultz]

If I look at your I.data structure, the size of the variables doesn't match:


fn: {68×1 cell}
Group: {116×1 cell}
Valence: {58×1 cell}
Arousal: {58×1 cell}
SS: {116×1 cell}

You have 68 input files, but 116 entries for Group and SS and then
only 58 for Group and Arousal. They should all be the sample length
(68 if there are 68 input files).

> --
> You received this message because you are subscribed to the Google Groups "Aaron's fMRI matlab tools." group.
> To unsubscribe from this group and stop receiving emails from it, send an email to fmri_matlab_to...@googlegroups.com.
> To post to this group, send email to fmri_mat...@googlegroups.com.
> To view this discussion on the web visit https://groups.google.com/d/msgid/fmri_matlab_tools/c2e68a37-e09a-42d1-b5e2-525d97726a07%40googlegroups.com.
> For more options, visit https://groups.google.com/d/optout.

[Haris Styliadis]

Hi Aaron, I have corrected the size of the variables for 116 input files.

Now I get the following error message

Error using char

Cell elements must be character arrays.

Error in ANOVA_APS (line 304)

    spacer = size(char([ModelTerms 'Error' posthocs(:,1)']),2)+4;

Error in GLM_Flex_Fast4 (line 52)

    MOD = ANOVA_APS(I.Data,I.Model,I.PostHocs,0,1,1);

Error in new_glm_flex (line 613)

GLM_Flex_Fast4(I);

Any ideas?  

Please have a look at the script (attached) and/or the workspace (attached) when you are available and guide me through tackling the error. Your advice on defining the post hocs will be highly appreciated.

Thank you for your time and support

Best,

Haris

> To unsubscribe from this group and stop receiving emails from it, send an email to fmri_mat...@googlegroups.com.