[FieldTrip] File Size depends on No. of Trials?

[Julian Keil]

Hi Christine,

take a look into the cfg.previous of ERDERS. Fieldtrip stores information
on previous data analysis steps there. Maybe the trial information is
hidden somewhere in there.
The MatLab command rmfield can then be used to remove some unwanted fields
from a structure. But please beware! There is a good reason for the
thorough bookkeeping in Fieldtrip (which has saved me quite often).
Good luck,

Julian

On Wed, Nov 30, 2016 at 10:25 AM, Blume Christine <christine.blume at sbg.ac.at
> wrote:

> Dear Community,
>
>
>
> I have an issue with data size. I am processing high-density EEG data from
> a sleep study. Following preprocessing I perform time-frequency
> transformations and a baseline correction using the following code:
>
>
>
>         cfg = [];
>
>         cfg.method = 'wavelet';
>
>         cfg.output = 'pow';
>
>         cfg.keeptrials = 'no';
>
>         cfg.width = 3; %
>
>         cfg.foi = 1:1:16; %
>
>         cfg.toi = -0.7:0.2:0.9;
>
>         freqanalysis_FV = ft_freqanalysis(cfg, data_FV);
>
>
>
>         cfg = [];
>
>         cfg.baseline = [-0.6, 0];
>
>         cfg.baselinetype = 'relchange';
>
>         ERDERS_FV = ft_freqbaseline(cfg, freqanalysis_FV)
>
>
>
> As you can see, I do *not* keep the trials. Still, the size of ERDERS
> seems to depend on the size of data_FV. Why is that the case? Also,
> changing cfg.foi does not change the file size – why? The problem is that
> this way the data can hardly be handled as it takes up so much RAM. Please
> note that I do clear everything from the workspace I do not need, that
> should not be the issue.
>
>
>
> Thanks a lot for your thoughts.
>
>
>
> Best,
>
> Christine
>
>
>
> _______________________________________________
> fieldtrip mailing list
> fieldtrip at donders.ru.nl
> https://mailman.science.ru.nl/mailman/listinfo/fieldtrip
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.science.ru.nl/pipermail/fieldtrip/attachments/20161130/0e8177d0/attachment.html>