[FieldTrip] preprocessing per channel

[Lieneke. Janssen]

Hi Eelke and others, 

Thanks for your reply. ft_appenddata might indeed solve my problem! 

It would be nice though if cfg.method would work the same for preprocessing per trial and preprocessing per channel. 
If cfg.method is not specified, fieldtrip assumes cfg.method = trial. It then preprocesses every trial subsequently and gives you one output file in which all subsequently preprocessed trials are present. 

Since I would like to preprocess my complete dataset as if one big trial as a first step, which apparently takes too much memory, I wanted to preprocess per channel (for all channels) instead of per trial. So I set cfg.method to channel. It then preprocesses every channel subsequently (as 'per trial'), but only gives you the output of the most recently processed channel (in contrary to 'per trial'). I therefore assume that output/data from the previously preprocessed channel gets lost (i.e. overwritten) when fieldtrip continues with the next channel. So unfortunately you end up with the output of the last preprocessed channel only. This does not happen for preprocessing per trial (i.e. ending up with the output of the last preprocessed trial only). 

I hope this problem can be solved sooner or later. Until then, I'll append my data :). 

Best, 
Lieneke  





----- "Eelke Spaak" <eelke.spaak at donders.ru.nl> wrote:

> From: "Eelke Spaak" <eelke.spaak at donders.ru.nl>
> To: "Email discussion list for the FieldTrip project" <fieldtrip at donders.ru.nl>
> Sent: Tuesday, April 26, 2011 11:45:52 AM
> Subject: Re: [FieldTrip] preprocessing per channel
>
> Dear Lieneke,
> 
> To be honest, I have never tried preprocessing per channel before, so
> I might not understand exactly what you mean by 'the output is
> overwritten with every channel that is preprocessed'. If so, please
> correct me :)
> 
> The cfg.method option governs only how data is read from disk, not
> which parts of the data the preprocessing is applied to. If you only
> want to preprocess certain channels, you can use cfg.channel = {...}
> to specify which channels to preprocess. If you have different data
> sets (e.g., corresponding to different output structures from
> ft_preprocessing) that contain different channels (but the same
> trials), you can concatenate them by using ft_appenddata
> (http://fieldtrip.fcdonders.nl/reference/ft_appenddata).
> 
> Hopefully this was somewhat helpful,
> Best,
> 
> Eelke
> 
> 2011/4/26 Lieneke. Janssen <lieneke.janssen at student.ru.nl>:
> > Dear Fieldtrippers,
> >
> > I was quite happy to find the possibility in fieldtrip to do
> preprocessing per channel instead of (the default) per trial.
> >
> > Unfortunately, if I say cfg.method = 'channel'; the output is
> overwritten with every channel that is preprocessed, so I end up only
> having the output from the last channel, whereas if the method is
> 'trial', the function nicely combines all preprocessed trials into the
> output file.
> >
> > Is there anything that can be done to get one preprocessing output
> file using cfg.method = 'channel';?
> >
> > Thanks!
> >
> > Best,
> > Lieneke
> > _______________________________________________
> > 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