[FieldTrip] Question about wavelet analysis

[jan-mathijs schoffelen]

Dear Inbal,

> 1. Just wanted to make sure:
> When I use [res] = ft_freqanalysis(cfg,data)  with data that have  
> several trials - and I define cfg.keeptrials = 'yes';
> I get an analysis for each trial in [res]. Then I use  
> ft_multiplotTFR(cfg,res) with cfg.trials = 'all';
> The result is an average on the wavelet result of all trials? (and  
> not a wavelet on the averaged signal of all trials)
>

Yes, ft_multiplotTFR computes an average spectrogram over observations/ 
trials

> 2. Is there a way for me to see the wavelet analysis result of each  
> trial and to browse easily between them?
>  rather than calling ft_multiplotTFR(cfg,res) each time with  
> cfg.trials = i ; where i is the desired trial number?
>

At the moment, no. I never really thought of looking at single trial  
multiplots but I can imagine the need for this functionality.
Your workaround indeed works, but may be a bit tedioius.

> 3.  How can I use other mother wavelet? (and not just Morlet wavelets)
>

What kind of wavelet would you like to use? If the wavelet consists of  
a tapered cosine/sine wave in which the Gaussian taper will be  
replaced by something else, you could use cfg.method  = 'mtmconvol'.  
This is an efficient implementation of a wavelet-like analysis, where  
there is some freedom to define your own taper. At present this relies  
on matlab's window-function (or on the slepian sequences; different  
story), so the different tapers are limited to those available in  
matlab. However, it is relatively straightforward to tap into the code  
here, and provide your own taper. A dangerous, but possible option is  
to temporarily overload the matlab window-function with your own,  
spitting out the requested taper.

> 4. How can I control the colormap min and max values so that they  
> will be similar in all figures? (like doing imagesc(pic,clim))?
>

I guess cfg.zlim should do the trick.

Best,

JM


> Thank you very much
>
> Inbal
> _______________________________________________
> fieldtrip mailing list
> fieldtrip at donders.ru.nl
> http://mailman.science.ru.nl/mailman/listinfo/fieldtrip

Dr. J.M. (Jan-Mathijs) Schoffelen
Donders Institute for Brain, Cognition and Behaviour,
Centre for Cognitive Neuroimaging,
Radboud University Nijmegen, The Netherlands
J.Schoffelen at donders.ru.nl
Telephone: 0031-24-3614793