[FieldTrip] ft_scalpcurrentdensity methods + error if method is not 'spline'

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

Hi Vitoria,
Could you first download a newer version of FieldTrip, e.g. 20150309 :-), and see if the problem is replicated with this version? I think it will persist, but it would be good to know for sure, otherwise we may be tackling problems that are not problems anymore.
Note that the low level function ft_datatype_sens throws an informative error, occurring on line 375. You may want to try to do a ‘dbstop if error’ to investigate it in a bit more detail. I suspect an ‘inconsistent number of channels in the sensor description’ :-). The reason why it may fail with the ‘spline’ method, and not with the other methods is probably that the former requires a description of the electrode positions, whereas the the latter methods only need a description of the neighbours for each electrode.

Best,
Jan-Mathijs

On Mar 10, 2015, at 8:56 PM, Vitoria Piai <v.piai.research at gmail.com<mailto:v.piai.research at gmail.com>> wrote:

Thank you, Cris.

What I noticed (but I fear I can't be more precise due to lack of information) is that you use two cfg definitions (cfg and cfgn) and that might create problems.
One of them (cfgn) is only being used to define neighbours, so I don't think that matters.
On top of that I would always clean the actual cfg by nulling it at the beginning of every function call (cfg=[]).
I always do this, only didn't paste that part of my code to have it occupying less space. So that isn't the problem either.
It would also help a lot if you included the call to ft_freqanalysis (again with an eye of regard of nulling the cfg beforehand).
I doubt it's the cfg in ft_freqanalysis that is causing the problem - it's the same cfg I've been using for 5 years now, and in line with the one shown in the tutorial. So unfortunately, not the solution either.

The key is in the computation of the scd depending on the method. The error only occurs with 'hjorth' and 'finite', but not with 'spline'. Maybe I should just read more on these three methods before I try to advance any further.

Thanks anyways, Vitoria




On Mon, Mar 9, 2015 at 4:58 AM, Vitória Piai <v.piai.research at gmail.com<mailto:v.piai.research at gmail.com>> wrote:
Hi all,

I'm interested in using Laplacian transformation prior to computing TFRs.
I still haven't read anything about the three methods implemented in FieldTrip, I must confess, but I think my question is independent of that. I'm using FT version: fieldtrip-20140801 on Matlab2014a.

If I use the method 'spline', I can subsequently run ft_freqanalysis (or ft_timelockanalysis for that matter). However, with either 'hjorth' or 'finite', I get an error when running ft_freqanalysis/ft_timelockanalysis:
My cfg:

  cfgn.method       = 'template';
  cfgn.layout       = 'biosemi64.lay';
  cfg.neighbours      = ft_prepare_neighbours(cfgn, dat);

  cfg.method = 'hjorth'; %'finite', 'spline'
  cfg.elec            = ft_read_sens('standard_1005.elc');
  lpc = ft_scalpcurrentdensity(cfg, dat);
% Then after that, a commonly used cfg for ft_freqanalysis

Error using ft_datatype_sens (line 375)
inconsistent number of channels in sensor description

Error in ft_datatype_raw (line 138)
      data.elec = ft_datatype_sens(data.elec);

Error in ft_checkdata (line 219)
  data = ft_datatype_raw(data, 'hassampleinfo', hassampleinfo);

Error in ft_freqanalysis (line 211)
data = ft_checkdata(data, 'datatype', {'raw', 'raw+comp', 'mvar'}, 'feedback', cfg.feedback, 'hassampleinfo', 'yes');

I understand that the "inconsistent number of channels in sensor description" is coming from these two other implementations, but should that be the case? The help on ft_scalpcurrentdensity says " The output data has the same format as the input and can be used in combination with most other FieldTrip functions". So I'm wondering whether there's something special intrinsic to these implementations, I'm using a too old FT version, my config isn't right, or this is an issue that has gone unnoticed because those implementations are not used often.

Thanks a lot!
Cheers from Berkeley,
Vitoria



_______________________________________________
fieldtrip mailing list
fieldtrip at donders.ru.nl<mailto:fieldtrip at donders.ru.nl>
http://mailman.science.ru.nl/mailman/listinfo/fieldtrip




_______________________________________________
fieldtrip mailing list
fieldtrip at donders.ru.nl<mailto:fieldtrip at donders.ru.nl>
http://mailman.science.ru.nl/mailman/listinfo/fieldtrip

_______________________________________________
fieldtrip mailing list
fieldtrip at donders.ru.nl<mailto:fieldtrip at donders.ru.nl>
http://mailman.science.ru.nl/mailman/listinfo/fieldtrip

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.science.ru.nl/pipermail/fieldtrip/attachments/20150310/1e904467/attachment-0002.html>