[Eeglablist] plot of components spectra and maps

Makoto Miyakoshi mmiyakoshi at ucsd.edu
Mon Jun 11 12:51:17 PDT 2012


Dear Ida,

> Does this mean that the calculation part was performed correctly all the
> time and that just the plot was not precise? To me it seems to be like that,
> since I am getting the same results, but with perfect plots, which I needed.

I think the default frequency resolution was just too coarse for your
purpose. You can change 'freqfac', 64 to other numbers to see how it
changes frequency resolution in your plot.

I'm glad you are glad.

Makoto

2012/6/11 ida miokovic <ida.miokovic at gmail.com>:
> Dear Makoto,
>
> this works perfectly! I did everything you said.
>
> Does this mean that the calculation part was performed correctly all the
> time and that just the plot was not precise? To me it seems to be like that,
> since I am getting the same results, but with perfect plots, which I needed.
>
> Thank you once again.
>
> Ida
>
>
> On Mon, Jun 11, 2012 at 6:50 PM, Makoto Miyakoshi <mmiyakoshi at ucsd.edu>
> wrote:
>>
>> Dear Ida,
>>
>> Ok, you should use the option 'freqfac'.
>> Follow the steps shown below.
>> 1. type 'edit' to open empty editor
>> 2. run the pop_spectopo from GUI
>> 3. type 'eegh' to get the log of the last command
>> 4. copy the pop_spectopo line of the log and paste it to empty editor.
>> 5. add this at the beginning:
>>       [spectra,freqs,speccomp,contrib,specstd] = pop_spectopo(...
>> 6. add 'freqfac', 64 at the end of the line
>> 7. select all lines in the editor and press F9
>> 8. open 'freqs' in the workspace. You should see a frequency
>> resolution of (0 to your nyquist)/16385 there.
>>
>> Let's say your sampling rate is 250Hz, and you are interested in
>> 10.1Hz spectrum and its map. Then, 10.1 is between 10.0937 and
>> 10.1013, and I think the latter is use for plotting.
>>
>> Makoto
>>
>> 2012/6/9 ida miokovic <ida.miokovic at gmail.com>:
>> > I just came to conclusion that these frequencies are rounded to the
>> > nearest
>> > value (0Hz, 0.5Hz, 1Hz) when plotting, am I right?
>> >
>> > Are they rounded in calculations also or a specified value is used? Is
>> > there
>> > a way for me to work with the actual frequency I want? How could I call
>> > this
>> > function as a code? I tried following:
>> >
>> > pop_spectopo('EEG',[-2000 1000],100,0.33)
>> >
>> > for current EEG data set, specified range, 100% of the data and 0.33 Hz
>> > as a
>> > frequency of interest, but something is not OK and I get following
>> > error:
>> >
>> > ??? Attempt to reference field of non-structure array.
>> >
>> > Error in ==> pop_spectopo at 120
>> > if ~isempty(EEG.chanlocs)
>> >
>> > Thank you very much for any help.
>> >
>> > Ida
>> >
>> >
>> > On Sat, Jun 9, 2012 at 3:15 PM, ida miokovic <ida.miokovic at gmail.com>
>> > wrote:
>> >>
>> >> Hello,
>> >>
>> >> I have question regarding procedure: Plot --> Components spectra and
>> >> maps.
>> >> I am interested in channel with the maximum power at the specified
>> >> frequency
>> >> so I put brackets [] in the electrode number to analyze and I use 100%
>> >> of
>> >> the data.
>> >>
>> >> Frequency of my interest is for example 0.27 Hz, but when the
>> >> calculation
>> >> is done, on the plot I get the scalps (components activities) connected
>> >> to
>> >> the frequency around 0.45 Hz (plotting frequency range was 0 1). Same
>> >> happens for the frequency 0.25 Hz (on the plot I get 0.45 Hz). On the
>> >> other
>> >> side, plot of the 1Hz is OK. In the workspace it is using value I put
>> >> in the
>> >> GUI, but on  the plot in case of 0.25, 0.27, 0.3, the frequencies are
>> >> not
>> >> those from the GUI.
>> >>
>> >> To me, it is important to get the channel of my interest and the
>> >> components where there is the most of specified frequencies. If it is
>> >> just
>> >> plotting issue, I don't mind, but any hints why does it happen would be
>> >> great. Maybe it is working better with integer values of frequencies...
>> >>
>> >> Thanks
>> >>
>> >> Ida
>> >
>> >
>> >
>> > _______________________________________________
>> > 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
>> JSPS Postdoctral Fellow for Research Abroad
>> Swartz Center for Computational Neuroscience
>> Institute for Neural Computation, University of California San Diego
>
>



-- 
Makoto Miyakoshi
JSPS Postdoctral Fellow for Research Abroad
Swartz Center for Computational Neuroscience
Institute for Neural Computation, University of California San Diego




More information about the eeglablist mailing list