[Eeglablist] Help about sift a plung in eeglab

[Tim Mullen]

Hi Tongxiao,

Well, the error message pretty much says it all. You have too many
parameters to fit and too few data samples. I presume that you have only
single-trial data and you are either trying to fit the multivariate model
to a very large number of channels/sources or your model order is very
large. The first thing you might do is try selecting fewer
channels/sources.

For more details on data limitations please see:

http://sccn.ucsd.edu/wiki/Chapter_6.6._Model_Fitting_and_Validation#6.6.1.3._Sufficient_amount_of_data

A second option is to use a regularized VAR model (e.g. ridge regression or
sparse VAR model fitting) which allows you to fit models with many more
parameters than data samples. Please contact me directly, if you want to
pursue this route.

Best,

Tim


On Fri, Oct 26, 2012 at 7:17 PM, <mengtongxiao at sina.com> wrote:

>
> Dear all,
>
> when I do the modle order selection,
> the command windows told me that,
>
> *EERORRatio of number of parameters to datapoints is 229.643.*
> * The ratio of number of parameters to datapoints must be <=1. *
> *your window length must be at least 114.822sec*
>
> But my data length is 96 sec.
> what should i do?
> thanks,
> Look forward,
> tongxiao
>
> _______________________________________________
> 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
>



-- 
---------  αντίληψη -----------
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://sccn.ucsd.edu/pipermail/eeglablist/attachments/20121031/b65c2f09/attachment.html>