Dear Muhammad,
You can make all your trials to have the same condition label and keep
the true labels in a separate list if you need them for stats. Then
the order will stay fixed.
Best,
Vladimir
On Thu, May 10, 2012 at 9:04 PM, Muhammad Adeel Parvaz
<[log in to unmask]> wrote:
> Hello SPMers,
>
> I am processing some EEG data for which I compute both a slow wave ERP and
> EEG alpha. I need to extract single-trial slow wave amplitude and EEG alpha
> power to do a within-subject (trial-by-trial) correlation between these
> variables for each condition. I am able to extract the single-trial slow
> wave amplitude, but having trouble with getting the EEG alpha power. After I
> convert the epoched data into alpha time-frequency domain (8-13 Hz), it
> seems that I need to perform 'convert2image' and average over 8 - 13hz
> frequencies to get the alpha time series. However, this processes
> 'rearranges' the epoch order and sorts them by condition type. This poses a
> problem for me because I need to extract the erp and alpha for each trial
> and both are now being extracted in different trial order. So, is there a
> way to either sort the epoched data by the condition or to prevent the
> convert2image function from sorting? Or, is there a better way to do this?
>
>
> Thanks
> -Muhammad
|