[FieldTrip] EGI mff file import

[Schoffelen, J.M. (Jan Mathijs)]

Hi Raphael,

If you are unsure about where the slowdown takes place you can evaluate this using the matlab profiler. If you type 'profview on’ a graphical interface to the profiler opens, and you can execute commands, which will then be stopwatched.

Best wishes,
Jan-Mathijs


On 28 Jun 2019, at 14:01, Raphaël Thézé <raphael.theze at gmail.com<mailto:raphael.theze at gmail.com>> wrote:

Dear Jan,

I have tried increasing the java heap memory but the processing still takes ~1 min per individual trial.

Regarding the code, I have tried to break down the different steps, and my impression was that the preproc function was slowing down the whole process, but this is only the case in the fieldtrip version with mff v3. I can try to go over the code again but the deeper I go into the FT functions, the more confusing it gets!

Thanks for the help,
Best
Raphaël

Le jeu. 13 juin 2019 à 18:27, Schoffelen, J.M. (Jan Mathijs) <jan.schoffelen at donders.ru.nl<mailto:jan.schoffelen at donders.ru.nl>> a écrit :
Dear Raphael,

I happened to bump into Trey Avery, from Philips/egi the other day, and he mentioned that you might want to consider to increase the java heap memory to a larger value. How this can be done is shown here: https://www.mathworks.com/help/matlab/matlab_external/java-heap-memory-preferences.html

That is, in your message you are a bit ambiguous as to where exactly in the code the slowdown takes place. If it is indeed preproc, as you mention, then it’s probably not an input/output problem.

Anyway, it wouldn’t hurt to try I think. Could you let us know whether this works for you? In that case we can add this to our documentation section of ‘getting started with egi data’ on our fieldtriptoolbox.org<http://fieldtriptoolbox.org/> website. Also, the current version of that part of the documentation could be improved a bit I think to reflect better the recent improvements to the code (i.e. the use of the v3 reader…). Feel free to suggest changes yourself.

Best wishes,

Jan-Mathijs

On 24 May 2019, at 15:58, Raphaël Thézé <raphael.theze at gmail.com<mailto:raphael.theze at gmail.com>> wrote:

Dear all,

I have a question regarding the import of mff file format.

I used to import my EGI data with fieldtrip version 20180201 which worked fine (with egi_mff_v1) but when I switched to fieldtrip version 20190325, it uses egi_mff_v3 as default. The latter also works well except that when it used to take 1 sec per trial to be imported with V1, now it takes about 1 minute per trial with V3. The amount of time needed to import one subject in my case would be theoretically 3 hours, which is ridiculously long. I understand that the switch of version takes place in ft_read_data, which is called by ft_preprocessing, but apparently it is preproc, also called by ft_preprocessing, that takes forever to compute, but I can't explain why. I tried to manually request to use the version egi_mff_V1 but then there is an error regarding xml file, that I have not tried to explore so far.

For now I will just go back to the 20180201 version of fieldtrip so I can quickly import data into matlab, but I would like to understand what is going wrong here, or if it is to be expected.

In advance, thank you all.
Best,
Raphaël
_______________________________________________
fieldtrip mailing list
https://mailman.science.ru.nl/mailman/listinfo/fieldtrip
https://doi.org/10.1371/journal.pcbi.1002202

_______________________________________________
fieldtrip mailing list
https://mailman.science.ru.nl/mailman/listinfo/fieldtrip
https://doi.org/10.1371/journal.pcbi.1002202
_______________________________________________
fieldtrip mailing list
https://mailman.science.ru.nl/mailman/listinfo/fieldtrip
https://doi.org/10.1371/journal.pcbi.1002202

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.science.ru.nl/pipermail/fieldtrip/attachments/20190629/931120a0/attachment-0002.html>
-------------- next part --------------
_______________________________________________
fieldtrip mailing list
https://mailman.science.ru.nl/mailman/listinfo/fieldtrip
https://doi.org/10.1371/journal.pcbi.1002202