[FieldTrip] ft_singleplotTFR seems broken

[Casper van Heck]

Dear Eelke,

I've put up an exerpt of my code, with the data from a single subject
(no.10, 9MB) in a separate folder
here<https://www.dropbox.com/sh/nr4s0yu4voh5w3l/6XvAvmaMdk>
.
It also includes a screenshot of what I see with when I run the code with
this exact dataset, and a small program 'filehandler', which I've been to
lazy to replace.

Note that this specific subject only has about 60 seconds of data (and I'm
still waiting for the student to tell me why).

I'm hoping you'll get the same result as the screenshot, and you'll be able
to tell me why...

Best regards,

Casper


On Wed, Sep 18, 2013 at 1:25 PM, Eelke Spaak <eelke.spaak at donders.ru.nl>wrote:

> Dear Casper,
>
> Could you maybe post a snippet online somewhere (e.g. dropbox) of data
> corresponding to this plot (i.e. the output of ft_freqanalysis that
> you input to ft_singleplotTFR, assuming that it's averaged over trials
> and therefore small)? Then we can see whether this is reproducable on
> our systems.
>
> Best,
> Eelke
>
> On 16 September 2013 17:12, Casper van Heck <caspervanheck at gmail.com>
> wrote:
> > Dear Fieldtrippers,
> >
> > I've been doing some exploratory plotting of a dataset where several
> > subjects were exposed to a continuous stimulus (CPM) over the course of
> > three minutes. One of the things that was asked was a time-frequency
> plot,
> > over those three minutes. While I'm not quite sure a TFR can be used on
> that
> > scale without problems, I went ahead with it just to see what Fieldtrip
> > would do. What I got, was this:
> > http://tinypic.com/r/zx2tqs/5|
> > And I have no idea what went wrong (note that this specific subject only
> got
> > a 60s stimulus). Quality of the picture and the validity of the analysis
> > aside, I have a severe problem with the axes, which seem to be all over
> the
> > place.
> >
> > I simply read the data using ft_preprocessing (no filters or anything,
> since
> > the student in charge already used a 10Hz high-pass and a 150Hz low-pass
> > filter).
> > The only field in 'cfg' is 'headerfile'.
> >
> > Next, I call ft_freqanalysis with:
> >     cfg.output = 'pow';
> >     cfg.method = 'wavelet'; % default; Morlet wavelet
> >     cfg.taper = 'hanning';
> >     cfg.tapsmofrq = 4;
> >     cfg.channel = 'Cz';
> >     cfg.foi = [20:0.5:30];
> >     cfg.t_ftimwin = ones(length(cfg.foi),1).*0.25;
> >     cfg.toi = 0:0.5:60;
> >
> > And finally, I end up with the monstrocity shown earlier. This problem
> only
> > presents itself with Fieldtrip-related functions, the normal Matlab-plots
> > work as well as they've ever done, so I'm thinking it's something in
> > Fieldtrip that's causing this.
> >
> > My hope is that someone has encountered this problem before, and knows
> how
> > to solve it...
> >
> > Sincerely,
> >
> > Casper
> >
> >
> > _______________________________________________
> > fieldtrip mailing list
> > fieldtrip at donders.ru.nl
> > http://mailman.science.ru.nl/mailman/listinfo/fieldtrip
> _______________________________________________
> fieldtrip mailing list
> fieldtrip at donders.ru.nl
> http://mailman.science.ru.nl/mailman/listinfo/fieldtrip
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.science.ru.nl/pipermail/fieldtrip/attachments/20130926/be5c5cdb/attachment-0002.html>