[FieldTrip] 'rule of thumb' for defining how many independent components to analyze for EEG?

[Eelke Spaak]

Dear Ellie,

In addition to Arnaud's comment; if your question is how much
components you should specify in your cfg.numcomponent parameter, I
would suggest to just leave it at the default ('all'). This will give
you as many components as their are channels in your data. Typically,
specifying anything less than that will just lead to a number of
components (the ones explaining least variance) being discarded.

Best,
Eelke

On 6 November 2012 19:44, Arnaud Delorme <arno at cerco.ups-tlse.fr> wrote:
> Dear Ellie,
>
> you should decompose your full rank matrix. PCA dimension reduction may seriously affects and bias ICA solutions (I have seen a yet-to-be published report on that).
>
> Best,
>
> Arno
>
> On 6 Nov 2012, at 08:31, Eleanor Harding wrote:
>
>> Dear Fieldtrip community,
>>
>> I intend to run ICA on an EEG dataset with 64 channels (+ 2 eye channels), on epochs which are about 4 seconds long, and I am unsure as to how many independent components I should define (20,65,..?).
>>
>> My question is, is there a 'rule of thumb' for defining how many independent components to analyze for EEG? In the archives I have found only references to MEG datasets with many channels, for example reducing over 200 channels to 80 components. Looking in the literature today I also wasn't able to interpret an answer.
>>
>> If anyone has any information, references, or other input I would be much obliged.
>>
>> Thanks,
>> Ellie
>>
>>
>>
>> --
>> ------------------------------------------------------------------
>> Eleanor Harding
>> PhD Student
>> Max Planck Institute for Human Cognitive and Brain Sciences
>> Stephanstraße 1A, 04103 Leipzig, Germany
>> Phone: +49 341 9940-2268
>> Fax: +49 341 9940 2260
>> http://www.cbs.mpg.de/~harding
>>
>>
>>
>> _______________________________________________
>> fieldtrip mailing list
>> fieldtrip at donders.ru.nl
>> http://mailman.science.ru.nl/mailman/listinfo/fieldtrip
>
>
> _______________________________________________
> fieldtrip mailing list
> fieldtrip at donders.ru.nl
> http://mailman.science.ru.nl/mailman/listinfo/fieldtrip