trialfun_neuromag

Robert Oostenveld r.oostenveld at FCDONDERS.RU.NL
Mon Jan 8 16:31:01 CET 2007


Dear Jyrki

On 8 Jan 2007, at 15:27, Thomas Witzel wrote:
> I was just going to add here as well. I think the
> trialfun_neuromag.m function could have a routine that searches
> first for STI101, if that is not there then for STI 014 and only if
> thats not there rely on user specified channels. This would
> increase the user comfort a bit.

I thought about it a little further, and now realise that I was
slightly too fast with my answer. You should probbaly not have used
trialfun_neuromag to begin with.

The trialfun_neuromag is mainly there for backward compatibility. It
is a very simple function, and therefore I do not object too much
against people using it. But my reason for not advocating it is that
it is not very convenient for conditional trial definition. The
read_fcdc_event function was written for that purpose, and it returns
an event structure that is similar for all data formats. Using that
event structure, it is much easier to make conditional trials (e.g.
stimulus should be "1", the  response should be "3" and the response
should be within 500ms after stimulus).

The preferred "default" trialfun to be used is trialfun_general. For
conditional trial definitions, you have to make your own trialfun.
That is documented on
http://www2.ru.nl/fcdonders/fieldtrip/doku.php?
id=fieldtrip:documentation:making_your_own_trialfun_for_conditional_tria
l_definition
The underlying idea implemented in read_fcdc_event and in the new
trialfuns (like "trialfun_general") is to seperate the "hardware
knowlegde" for getting trigger information out of the file from the
"experimental knowledge" in determining the sequence of trigger
events that is interesting.

Looking in read_fcdc_event.m  (->private/read_event.m), I see that
the suggestion by Thomas is already in there more or less (i.e.
searching for the known trigger channels). The analog noise handling
there is still not optimal. Suggestions for improvements in
read_event are welcome.

best regards,
Robert

PS I will add this information to the documentation of all backward
compatibility trialfuns that are in private, probably I will also add
a warning there to discourage people from using them.



More information about the fieldtrip mailing list