Hi Debbie,
On Mon, Mar 28, 2011 at 12:54 PM, Deborah Talmi
<[log in to unmask]> wrote:
> Hi,
>
> I'm averaging a noisy dataset. Regular averaging works fine, but when I choose 'robust averaging' the following error appears when I try to display the robust averaged dataset.
>
>
> SPM8: spm_eeg_average (v3497) 12:40:42 - 28/03/2011
> ========================================================================
> Robust averaging finished after 14 iterations.
> macefdspm8_P06LB_p01.mat: Number of replications per contrast:
> average 1251: 30 trials, average 1751: 90 trials, average 251: 30 trials, average 751: 91 trials, average 1250: 30 trials, average 250: 30 trials, average 1750: 89 trials, average 750: 90 trials, average 125: 60 trials, average 175: 179 trials, average 25: 60 trials, average 75: 181 trials.
> Robust averaging might have introduced high frequencies in the data. It is advised to re-apply low-pass filter
> ??? Error using ==> set
> Bad value for axes property: 'YLim'
> Values must be increasing and non-NaN.
>
> Error in ==> spm_DisplayTimeSeries at 338
> set(ud.v.handles.gpa,...
>
> Error in ==> spm_eeg_review_callbacks>updateDisp at 994
> ud = spm_DisplayTimeSeries(D.data.y(:,:,trN(1))',options);
>
> Error in ==> spm_eeg_review_callbacks at 723
> updateDisp(D,1)
>
> ??? Error while evaluating uicontrol Callback
>
> Help would be appreciated.
>
This is something in the data (most likely NaNs). Robust averaging
should be robust to NaNs but if somehow you have a channel with
nothing but NaNs it probably would still output them. I can't tell you
what's wrong just from the error.
> I also noticed that spm_eeg_sort_conditions only works on the merged dataset prior to averaging (when this sorted file is averaged it produces the correct condition order) but doesn't appear to change the order of conditions when used on the averaged dataset. Is this intended?
>
Sort conditions does not physically change the order of the trials but
it does change the order in which they are processed by SPM functions
(like source reconstruction and DCM). If you sort conditions before
averaging then the average dataset will be physically sorted in that
order because the averaging functions goes through the conditions in
the order of D.condlist . If you sort an averaged dataset the data
file will stay the same but if you do source reconstruction what is
meant by condition 1, condition 2 etc. will be modified.
Best,
Vladimir
>
> Cheers,
> Deborah
>
|