[Eeglablist] Spectopo Function Negative Power Values

Matt Craddock m.p.craddock at leeds.ac.uk
Mon Nov 13 10:39:36 PST 2017


Hi,

Note also that there is nothing wrong with spectopo producing negative 
power values. The values it returns are decibels (as the y-axis on the 
plots says, "Log Power Spectral Density 10*log10(muV^2/Hz)"). So the 
question is really whether the negative values are overly "large". If 
you really wanted you could transform it back to just being muV^2/Hz 
using 10 ^ (spectrum / 10).

Cheers,
Matt

On 10/11/2017 17:47, Tarik S Bel-Bahar wrote:
> Note that you can also look in the spectopo output to see if the values 
> are negative or not when initially computed by the function.
> 
> Note also that the graphical output for spectopo may transform the 
> spectral data, so you also want to try plotting the output data directly.
> 
> Thanks for letting the list know when you have a solution so that other 
> users can benefit from your experience.
> 
> Last, see Chronux toolbox and also Mike X. Cohen's time frequency book/code.
> 
> 
> On Nov 9, 2017 10:24 AM, "Tarik S Bel-Bahar" <tarikbelbahar at gmail.com 
> <mailto:tarikbelbahar at gmail.com>> wrote:
> 
>     ​Hello Courtney,
> 
>     *If you haven't yet, it would be best to do all steps with eeglab
>     tutorial data, and see if you get normal values following the
>     recommended steps in the eeglab tutorial. Use the GUI only, and
>     examine the eegh output after successful steps. Do you see similar
>     large negative values ? Is your code different from the eegh output ?
> 
>     *Also, do you see similar negative values across all your
>     single-subject files or just some? If just some, then this might be
>     an issue with some of your single subjects.
> 
>     *Make sure you are doing preprocessing steps correctly and in right
>     order. How you filter and clean the data is important.
> 
>     *If you reach a solution please share with the list so that other
>     users can benefit from your experience.
> 
>     *You may also benefit from reading Mike X. Cohen's excellent
>     book/excercises/videos about time-frequency for eeg.
> 
>     ​*In the future, it would be beneficial to send pictures and more
>     info about your previous steps. that will make it easier to better
>     understand your issues and possibly get more responses.
> 
> 
> 
>     On Tue, Nov 7, 2017 at 7:11 AM, Courtney Jimenez
>     <cojimenez at ucdavis.edu <mailto:cojimenez at ucdavis.edu>> wrote:
> 
>         Hi all,
> 
>         We are using eeglab's spectopo function to calculate power
>         spectral density estimates for our data. However, we're getting
>         large negative values in many channels across many frequencies.
> 
>         How can we correct this?
> 
>         We use the following code from the matlab command line:
>         spectopo(EEG.data(:,:),0,EEG.srate,'freqrange',[4
>         50],'title',EEG.setname)
> 
> 
>         Thanks,
>         Courtney
> 
>         _______________________________________________
>         Eeglablist page: http://sccn.ucsd.edu/eeglab/eeglabmail.html
>         <http://sccn.ucsd.edu/eeglab/eeglabmail.html>
>         To unsubscribe, send an empty email to
>         eeglablist-unsubscribe at sccn.ucsd.edu
>         <mailto: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
>         <mailto:eeglablist-request at sccn.ucsd.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
> 


More information about the eeglablist mailing list