[FieldTrip] Logarithmic plot of power spectrum

[Joram van Driel]

Hi Daniel,

Alternatively, you could use contourf instead of the default fieldtrip
plotting function. This function has more flexibility for rescaling of axes
than the imagesc function. From the ft_freqanalysis result, extract a
time-by-frequency matrix, and then simply make your own figure:

figure;
contourf(cfg.toi, cfg.foi,
TFR.powspctrm(chanidx,:,:),40,'linestyle','none');
set(gca,'yscale','log','ytick',[2 4 8 16 32 64]);

The chanidx is the channel you want to plot; the value 40 is the smoothness
of the contours, and here it is plotted without lines between the contours
(default is 'yes'). The set(gca... part changes the y-axis scaling,
irrespective of whether your frequencies were logarithmically spaced in
your analysis.

Chrs,
Joram

On Mon, Jul 6, 2015 at 6:02 PM, Daniel Hähnke <
daniel.haehnke at lrz.tu-muenchen.de> wrote:

> Dear Philip,
>
> you’re right, my question is about the equal spacing of the y-axis.
>
> Thanks for the suggestion! I was just using the same line of thought to
> change the code of ft_singleplotTFR, but your approach is much safer than
> messing around with fieldtrip code ;)
>
> All the best,
>
> Daniel
>
> On 06 Jul 2015, at 17:54, Philipp Ruhnau <mail at philipp-ruhnau.de> wrote:
>
> Dear Daniel,
> you can plot TFRs with logarithmic axes yes, by just using ft_multiplotTFR
> or ft_singleplotTFR with your logspace TF data
> but I think you’re question is about equally spacing the image?
>
> if so you can ’trick’ fieldtrip, or better matlabs imagesc which fieldtrip
> uses to plot, by changing the values in the data.freq field to a linearly
> increasing vector (1:1:numel(data.freq)) and then
> either change the numeric values back in a graphic design software,
> or you change the yticklabels in matlab. see snippets below (just tested
> this quickly so no guarantee)
> hope this helps
> philipp
>
>
>
> %assume you have a dataset 'data' which has a powspectrum field
> % keep the original frequency axis
> orig_freq_axis = data.freq;
> % now create fake linear axis
> data.freq = 1:numel(data.freq);
>
> % plot data with defaults (needs powspectrum)
> cfg = [];
> cfg.channel = 1:10; % average across first 10 channels
> figure; ft_singleplotTFR(cfg, data)
>
> % get the y tick labels, which we can use as index for the original data
> yLabels = get(gca, 'YTickLabel');
>
> % find index and create new labels in a cell of strings
> newYLabels = [];
> for i=1:numel(yLabels)
>   ind = str2num(yLabels{i});
>   newYLabels{i} = num2str(orig_freq_axis(ind));
> end
>
> % now change the current labels to correct labels
> set(gca, 'YTickLabel', newYLabels')
>
>
>
> On 06 Jul 2015, at 16:09, Daniel Hähnke <daniel.haehnke at lrz.tu-muenchen.de>
> wrote:
>
> Dear community,
>
> I feel like this is a very basic question, but I have not found any
> solution here.
>
> I’d like to plot my TFRs with a logarithmic frequency axis, as seen e.g.
> in http://www.pnas.org/content/106/50/21341.figures-only
>
> Is there any way I can do this with fieldtrip? I know that I can
> logarithmically space the frequencies of interest, so this question is
> merely about plotting.
>
> All the best,
>
> Daniel
> _______________________________________________
> fieldtrip mailing list
> fieldtrip at donders.ru.nl
> http://mailman.science.ru.nl/mailman/listinfo/fieldtrip
>
>
>
>
> --------------------------
> Philipp Ruhnau, PhD
> CIMeC - Center for Mind/Brain Sciences
> Università degli Studi di Trento
> via delle Regole, 101
> 38123 Mattarello (TN) - ITALY
>
> phone: +39 0461 28 3083
> fax: +39 0461 28 3066
> homepage: http://sites.google.com/site/obobcimec
>
>
> _______________________________________________
> 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
>



-- 
Joram van Driel, PhD
Postdoc @ Vrije Universiteit Amsterdam
Cognitive Psychology
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.science.ru.nl/pipermail/fieldtrip/attachments/20150706/17e0b7a3/attachment-0002.html>