eye(5) would indeed do the trick! For "any effect" F-contrast one can also use the "specify columns for null space" which in your example would be 6:12 This option is of particular interest when the columns of interest are between other columns of no interest.
Hope this helps
Torben
Sendt fra min iPhone
> Den 27/08/2015 kl. 00.22 skrev briannh <[log in to unmask]>:
>
> Isn't just eye(5) sufficient??
> Best, Brian
>
>
>> On 2015-08-26 18:46, Dr Cyril Pernet wrote:
>> Sure was 😃
>>
>> Sent from my HTC
>>
>> ----- Reply message -----
>> From: "MCLAREN, Donald" <[log in to unmask]>
>> To: <[log in to unmask]>
>> Subject: [SPM] effect of interest F contrast
>> Date: Wed, Aug 26, 2015 16:41
>>
>> I think there was a missing , in the zeros command: [eye(5)
>> zeros(5,7)]
>>
>> Best Regards,
>> Donald McLaren, PhD
>>
>> On Wed, Aug 26, 2015 at 11:17 AM, cyril pernet <[log in to unmask]
>> [1]> wrote:
>>
>>> try [eye(5) zeros(5 7)] that should do the trick
>>> ie one column per regressor of interest
>>>
>>> cyril
>>>
>>>> Dear experts,
>>>>
>>>> I like to extract time series from different regions and adjust
>>>> this by the effect of interest F contrast. In my design matrix I
>>>> have 5 regressors of interests and the 6 movement parameters. Can
>>>> you please help how to define the effect of interest F contrast
>>>> for this case?
>>>>
>>>> Thank you very much
>>>> Best wishes
>>>> André
>>>
>>> --
>>> Dr Cyril Pernet,
>>> Centre for Clinical Brain Sciences (CCBS)
>>> Neuroimaging Sciences
>>>
>>> The University of Edinburgh
>>> Chancellor's Building
>>> Room GU426D
>>> 49 Little France Crescent
>>> Edinburgh EH16 4SB
>>>
>>> [log in to unmask]
>>> tel: 0131 465 9530
>>> http://www.sbirc.ed.ac.uk/cyril
>>> http://www.ed.ac.uk/edinburgh-imaging
>>>
>>> The University of Edinburgh is a charitable body, registered in
>>> Scotland, with registration number SC005336.
>>
>>
>>
>> Links:
>> ------
>> [1] mailto:[log in to unmask]
|