Dear Boris,
this is indeed the 'new' DCM10 structure. What you are looking for can
be found in structure DCM.Ep containing the conditional means of the
parameters of the model.
Best regards,
Guillaume.
On 13/09/10 09:25, boris suchan ruhr universität wrote:
> Hi Stephen,
> whne loading the DCM file I have something like:
>
>
> DCM =
>
> a: [4x4 double]
> b: [4x4x3 double]
> c: [4x3 double]
> d: [4x4x0 double]
> U: [1x1 struct]
> Y: [1x1 struct]
> v: 210
> n: 4
> TE: 0.0400
> delays: [4x1 double]
> options: [1x1 struct]
> M: [1x1 struct]
> Ce: [4x1 double]
> Ep: [1x1 struct]
> Cp: [85x85 double]
> Pp: [1x1 struct]
> Vp: [1x1 struct]
> H1: [32x4x3 double]
> K1: [32x4x3 double]
> R: [210x4 double]
> y: [210x4 double]
> T: 0
> F: -1.0783e+003
> ID: 414.6879
> AIC: 13.0189
> BIC: -42.2084
> xY: [1x4 struct]
>
>
>
>
> DCM.a seems for me to be the definition of the driving inputs and
> intrinsic connection...
>
>
> cheers
>
> boris
>
>
> Am 10.09.2010 17:39, schrieb Fromm, Stephen (NIH/NIMH) [C]:
>> I checked the newest version of SPM that I have, 4010, which I think is the newest available version, and I still see the matrices there.
>>
>> Stephen J. Fromm, PhD
>> Contractor, NIMH/MAP
>> (301) 451--9265
>> ________________________________________
>> From: Boris Suchan [[log in to unmask]]
>> Sent: Friday, September 10, 2010 11:20 AM
>> To: Fromm, Stephen (NIH/NIMH) [C]; [log in to unmask]
>> Subject: Re: DCM in SPM8 A and B matrices
>>
>> Dear all,
>> this is the way the "old" spm8 version worked the newest version has a
>> different structure and nothing is comparable to this structue.
>> I am sorry but does anybody has an idea?
>> best wishes
>> boris
>>
>>
>> Stephen J. Fromm schrieb am 2010-09-10:
>>> Load the DCM*.mat file into matlab. (It's enough to have specified
>>> the DCM model for a subject.) The a, b, and c matrices are fields of
>>> the DCM structure. E.g.:
>>>>> load DCM_myDCM
>>>>> whos
>>> Name Size Bytes Class
>>> DCM 1x1 1400344 struct array
>>> Grand total is 162999 elements using 1400344 bytes
>>>>> DCM
>>> DCM =
>>> a: [8x8 double]
>>> b: [8x8x4 double]
>>> c: [8x4 double]
>>> U: [1x1 struct]
>>> Y: [1x1 struct]
>>> xY: [1x8 struct]
>>> v: 185
>>> n: 8
>>> delays: [8x1 double]
>>> TE: 0.0230
>
>
--
Guillaume Flandin, PhD
Wellcome Trust Centre for Neuroimaging
University College London
12 Queen Square
London WC1N 3BG
|