Print

Print


Dear Christina and Janina
My apologies for the delay in replying. The procedure you describe sounds perfectly normal - I have not come across this problem before. Can I check - is this a recently downloaded version of SPM? And if you delete the BMS.mat file and re-run the batch, does the same problem occur again? If so, please could you copy and paste the output from the main Matlab window, which is produced when you run the batch?

All the best
Peter

-----Original Message-----
From: SPM (Statistical Parametric Mapping) [mailto:[log in to unmask]] On Behalf Of Janina Seubert
Sent: 04 April 2017 13:47
To: [log in to unmask]
Subject: [SPM] empty cell for BMA?

Hi,
I am trying to run BMA on 5 DCMs following the instructions in the manual, but it's not working and I am wondering if anyone might be able to give me a hint on what we are doing wrong. 
In the model inference batch, I put in the five DCM matfiles per subject, selecting random effects and leaving family as "Load family". I then choose "winning family" for BMA. When I then load the resulting BMS.mat file, however, the BMS.DCM.rfx.bma is empty.

BMS.DCM.rfx

ans = 

       data: '/Volumes/K8_Shared$/Server/KIPROJECTS/Q8_KI_fMRI_analysis/Subjects/DCM_group_BMS_rfx/mod...'
    F_fname: []
          F: [18x5 double]
         SF: [-2.0957e+05 -2.0951e+05 -2.0816e+05 -2.0772e+05 -2.0817e+05]
      model: [1x1 struct]
     family: []
        bma: []

BMS.DCM.rfx.model looks fine to me, as far as I can tell.

ans = 

    g_post: [18x5 double]
     alpha: [0.9943 2.0141 5.0742 9.9644 4.9529]
     exp_r: [0.0432 0.0876 0.2206 0.4332 0.2153]
        xp: [1.0000e-04 0.0035 0.0820 0.8341 0.0803]
       bor: 0.0596
       pxp: [0.0120 0.0152 0.0890 0.7963 0.0874]

Any advice would be very much appreciated! Thanks, Janina