[Eeglablist] Is there an inherent matlab version of newtimef

[Armand Mensen]

Dear EEGLAB users,

In my efforts to both understand, and also analyse a particular dataset I'm
working on, I'd like to get as close to the output for the ersp of the
"newtimef" function using the matlab "cwt" function.

I'm getting pretty close by just using:
        trial_wavelet(:, :, nT) = cwt(...
            double(EEG.data(nCh, :, nT)), ...
            'amor', EEG.srate);

And then identical normalisation parameters that are in newtimef (baseline
correction etc).

However its not quite right. In particular it seems that higher frequencies
are not quite given the same treatment. Are there settings etc that would
basically reproduce the newtimef function?

As a background: The idea of high-frequency suppression by a slow wave in
sleep is interesting, and when I explore the ERSP using newtimef I see the
slow power increase followed by high frequency suppression of equal
(normalised) magnitude. However with the cwt function I see the slow power,
but there is basically no decrease in the higher range.

Any help here would be greatly appreciated.

Take care,
Armand Mensen
Postdoc Researcher
University Hospital Bern, Switzerland
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://sccn.ucsd.edu/pipermail/eeglablist/attachments/20180312/50e99d7b/attachment.html>