JISCMail - SPM Archives

Dear Qasim

I have CC’d the SPM mailing list – you sent this to the FSL mailing list in error. There were three problems with the data you sent:

1. There was a bug in the SPM batch, when checking that all models within a subject were fitted to the same timeseries. I have now fixed this for future SPM releases. In the meantime, please set ‘Verify data identity’ to No in the batch.

2. You have set up the batch to have 3 models for the same subject. However, each of your models is for a different subject. You need to click ‘New:subject’ twice more in the batch. Note that you need more than one DCM per subject to perform model comparison.

3. I also noticed that your timeseries look strange. Take a look by typing:

load('DCM_Subj1.mat');

figure;plot(DCM.Y.y);

You’ve got big spikes at the start and small fluctuations after that.

Best

Peter

From: Peter Z
Sent: 21 September 2016 11:44
To: Zeidman, Peter <[log in to unmask]>
Subject: Fwd: BMS error: Reference to non-existent field 'y'

---------- Forwarded message ----------

Dear SPM experts,

I have already estimated the DCMs and now trying to run BMS/FFX to select the final model, however I am getting the following error

Failed 'Model Inference'

Reference to non-existent field 'y'.

I have attached my DCMs that I m inputting, these DCMs are from three different subjects. I have used a fully connected model. Can someone please check if these DCMs are correctly estimated ?

I have also attached my BMS code

Any suggestions to remove the error will be very useful

Thanks a lot

kind regards,

Qasim