Dear Peter,

Thank you so much for your help. I will check the outcomes carefully.

Best,
Andy

On Mon, Jan 11, 2016 at 7:55 PM, Zeidman, Peter <[log in to unmask]> wrote:
Dear Andy (CC'd mailing list),
You asked for a script that would create DCMs where some connections are the same across models, some are always absent, and some vary across models. I attach an updated version of your previous script to do this. Within the script, you can define which connections stay the same or differ according to the matrix:

% 0 = no modulation, 1 = always modulated, 2 = vary with/without modulation
M = [2 0 2;
     1 2 2;
     2 2 2];

You also asked for a specific rule that models should not be included if there are no modulatory inputs between regions R1 and R3. I have included code for this on lines 87-90.

You mentioned you wish to assign each model to a family. I attach a function to do this. It creates a file called families.mat. When you create the BMS, select "Load family" under "Family inference", and select this file. NOTE: You must select the models in the same order for the script as the order you select for the BMS batch.

Please thoroughly check the DCMs and the families.mat which are created, to ensure I have understood your request correctly and not made any mistakes. When I tested it, it produced 96 models with 32 assigned to each family.

Best,
Peter