[Eeglablist] ICA decomposition filtering and subsampling

[Julie Onton]

Hi Marco,

More points are helpful when they are, indeed, more data (ie, more time).
Having a higher sampling rate might have minor advantages for relatively messy
datasets, but for good clean data, a higher or lower sampling rate probably
makes no difference. However, I have not tried decompositions with less than
250 Hz sampling rate, nor over 512Hz, so all I can say is that the difference
between 256 and 512 is generally hard to detect.

Best, Julie

-- 
Julie Onton, PhD
http://sccn.ucsd.edu/~julie

> Hi everyone,
>
>
>
> I'm a new EEGLab user and I have a (probably trivial) question about ICA
> decomposition. I have a dataset with a sampling frequency of 2048 Hz. Then,
> for my analysis I filtered EEG data considering only the [0.1 60] Hz band. I
> read that it is better to use in the ICA decomposition as many points as
> possible but conidering the filters that I used and the sampling theorem a
> 120 Hz sampling frequency should be enough. After filtering, should I keep
> the original sampling frequency to improve the component detection or is it
> possible a sub-sampling?
>
>
>
> Thanks so much
>
>
>
> Greetings.
>
> Marco
>
> --
>
> Marco Bellami
>
> PhD Student
>
> Università la Sapienza
>
> Roma, Italy
> _______________________________________________
> Eeglablist page: http://sccn.ucsd.edu/eeglab/eeglabmail.html
> To unsubscribe, send an empty email to eeglablist-unsubscribe at sccn.ucsd.edu
> For digest mode, send an email with the subject "set digest mime" to
> eeglablist-request at sccn.ucsd.edu