[Eeglablist] SIFT seems to be stuck using a single data point, not whole epoch - help!

Makoto Miyakoshi mmiyakoshi at ucsd.edu
Mon Aug 11 12:34:56 PDT 2014

Dear James,

Just yesterday Tim Mullen, the developer of SIFT, noticed the error in
distributing the latest SIFT through the EEGLAB plugin manager, so that may
be related... this will be fixed (hopefully) very quickly.

but I want to say that I've heard the same thing before, and the cause of
the problem at that time was that the user used the default values for
preprocess, which was like 'for 2-sec epoch, use 2-sec sliding window' in
which the window can not slide at all and could generate indeed a single
point (like FFT, in contrast to wavelet transform). In your case the
message says '6/105' so this may not be the case, but double check it just
in case.

I also recommend that you check EEG.CAT for the parameters of ongoing SIFT
analysis because the PropertyGrid tool sometimes behaves counterintuitively
and users may find the different parameters in the computation from what
they specified.


On Tue, Aug 5, 2014 at 3:39 PM, James Jones-Rounds <jj324 at cornell.edu>

> Hello all,
> While using the newer versions of SIFT (1.33), EEGLAB (13.3.2b), and
> MATLAB (2014a) with our data, we're noticing strange behavior that seems to
> suggest the SIFT processing steps are only using a single data point
> instead of the entire epoch.
> The preprocessing steps appear to work, although rather quickly, and do
> not give the typical status update indicating which window is being
> analyzed at that moment (status update just says "checking whiteness,
> Condition: " instead of "checking whiteness, Condition: , windows 6/105",
> etc.). These steps complete much more quickly than usual. Our concern is
> that a single data point is being used as opposed to the whole epoch.
> Additionally, when using the model order selection step, the model order
> selection assistant does not give a resulting line plot indicating how many
> windows exceed or don't exceed the thresholds for whiteness or consistency.
> There's just a single bar graph for one of the whiteness evaluating
> algorithms, and the rest are blank. The visualization of connectivity also
> produces a familiar component vs. component matrix of graphs, but all of
> them are empty. Finally, when we go to use the BrainMovie feature, only a
> single time-point seems to be available in the bottom time-scroll bar (i.e.
> the "start" time point and the "end" time point are the same time point, in
> our case, -1.88 seconds).
> This appears to be the case for other datasets we're trying, and we're
> also using the recommended steps found in the SIFT manual. How would we go
> about verifying whether a single data point in the epoch is being used, and
> has anyone else noticed this issue? Is it a MATLAB-EEGLAB-SIFT version
> compatibility issue?
> Thank you all for your attention!
> James
> --
> James Jones-Rounds
> Laboratory Manager
> Human Development EEG and Psychophysiology (HEP) Laboratory,
> Department of Human Development,
> --------------------------------------------
> Cornell University | Ithaca, NY
> 607-255-9883
> eeg at cornell.edu
> _______________________________________________
> 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

Makoto Miyakoshi
Swartz Center for Computational Neuroscience
Institute for Neural Computation, University of California San Diego
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://sccn.ucsd.edu/pipermail/eeglablist/attachments/20140811/cf87c7e9/attachment.html>

More information about the eeglablist mailing list