Overlapping trials

Tue Jul 8 16:52:34 CEST 2008

Dear Natalia,

Most likely your error is caused by the begsmp/endsmp in read_data
not being integer and positive. This could be due to the fact that
your variables nlfpsamplesprestimbeamcrossing and
triallengthinlfpsamples are not integer numbers. A simple round
should do the trick I guess...
In general, I'd advice you to use less elaborate names for the
definition of your variables: it saves some typing (also for the
repliers to your question ;o) ), and so reduces the probability of
typos. Moreover, it facilitates reading.

Hope this helped,

Jan-Mathijs

On Jul 8, 2008, at 4:56 PM, Natalia Grion wrote:

> Hi Thomas and All,
> well this are recordings from freely moving rats with neuralynx
> system. By the way I'm defining trials, yes, they have variable
> length: I'm taking 3sec before the central beam is triggered (this
> is indicative of stimulus "on" = start trigger), till the trigger
> of the 2nd beam is trigger (reward trigger), even if they have a
> regular behaviour from trial to trial, sometimes they are faster.
> Start trigger is stored in a separate file as EventTimeTrigger,
> this event is a TimeStamp, Reward Trigger is stored in an external
> file (txt file) generated by labview (software that commands the
> maze). Timestamps are generated by the recording system
> (microseconds).  Every LFP sample has an associated timestamp and
> this forms the time axis,  knowing the EventTimeStamp i can take
> the 'n' samples before this event and/or after, in my case time
> taken 'after' depends on how long the rat took to trigger the
> Reward beam  (ellapsed time). This is my function (main part):
>
> function [trl]=trialfunNG(cfg)
> *% get the inter-timestamp time from the header*
> hdr=read_fcdc_header(cfg.datadir);
> nlfpsamplespersec=hdr.Fs;
>
> *% read the timestamps of the events in the Neuralynx stimulus
> channel (the .nev-file in cfg.datadir)*
> events=read_fcdc_event(cfg.datadir);
> eventlfpsamples=[events(:).sample]';
>
> *% read the stimulus and behavioral information in the external
> file....info about stimulus, correct\incorrect trials etc*
> stimbehdata=dlmread(cfg.extstimbehfile,'\t',1,1);
> *....*
> nlfpsamplesprestimbeamcrossing=cfg.prestimbeam*nlfpsamplespersec;
>
> if ischar(cfg.poststimbeam) && strcmp(cfg.poststimbeam,'respbeam')
> *%this respbeam is taken from the txt file*
>  rtinmsec=stimbehdata(selvec,6);
>  triallengthinlfpsamples=(rtinmsec/1000)*nlfpsamplespersec;
> else
>  triallengthinlfpsamples=cfg.poststimbeam*nlfpsamplespersec;
> end;
> trl=zeros(nseltrials,3);
> trl(:,1)=eventlfpsamples(selvec)-nlfpsamplesprestimbeamcrossing;
> trl(:,2)=eventlfpsamples(selvec)+triallengthinlfpsamples;
> trl(:,3)=-nlfpsamplesprestimbeamcrossing;
>
> I don't know till this point if I'm missing sth from Nlx, or is a
> bug generated by the way fieldtrip is picking up my data. One
> solution I thought,  would be to write a conditional statement for
> skipping these trials, but, as the error comes up from read_data.
> I'd like to know how to introduce it.
> Thanks again
> Natalia
>
>
>
> Thomas Dijkman wrote:
>> Hi Natalia,
>>
>> Your error seems te originate from this line: dat = ncs.dat
>> (begsample:endsample); Either the begsample variable or the
>> endsample variable isn't a positive number. As you probably know,
>> matlab starts counting with 1 for vector and matrix indices. Since
>> you mention the error doesn't come up when you take a trial 2 sec
>> instead of 3 sec before the prestimbeam, my guess is that he
>> sample index of the prestimbeam is more than 2 sec * samplefreq,
>> but less then 3 sec * samplefreq. Perhaps you could elaborate a
>> bit further about your experimental setup? All your trials have a
>> different duration? I hope this helps,
>>
>> Thomas Dijkman
>>
>> -----Oorspronkelijk bericht-----
>> Van: FieldTrip discussion list namens Natalia Grion
>> Verzonden: di 8-7-2008 11:50
>> Aan: FIELDTRIP at NIC.SURFNET.NL
>> Onderwerp: [FIELDTRIP] Overlapping trials
>>  Dear All,
>> During reading and preprocessing lfps,  I find this error for some
>> trials:
>>
>> ?? Subscript indices must either be real positive integers or
>> logicals.
>> *Error in ==> read_data at 576
>>            dat = ncs.dat(begsample:endsample);*
>> Error in ==> read_fcdc_data at 49
>> [dat] = read_data(varargin{:});
>> Error in ==> preprocessing at 554
>>     dat = read_fcdc_data(cfg.datafile, hdr, begsample, endsample,
>> rawindx, strcmp(cfg.continuous,
>>     'yes'));
>> Error in ==> FreqAnalysis at 51
>>   datar{chanindx}=preprocessing(cfgpreproc)
>>
>> Trials are defined by 2 trigger beams, first trigger defines the
>> start of the trial, and 2nd trigger beam: the reward time (which
>> is variable). I'm taking smaples in this way: prestimbeam: 3 sec
>> before "start trigger", and poststimbeam = time till trigger of
>> reward. I pretty sure this error appears when prestimbeam is
>> overlapping with the  previous poststimbeam (actually, taking
>> 2sec before start trigger error doesn't come up).   Do you have
>> any suggestion of how i could introduce i conditional statement
>> for skipping these trials?  I guess this statement has to be
>> written is "read_data" function, right?
>> Thanks!
>> Natalia Grion
>>
>> ----------------------------------
>> The aim of this list is to facilitate the discussion between users
>> of the FieldTrip  toolbox, to share experiences and to discuss new
>> ideas for MEG and EEG analysis. See also http://
>> listserv.surfnet.nl/archives/fieldtrip.html and http://www.ru.nl/
>> fcdonders/fieldtrip.
>>
>> ----------------------------------
>> The aim of this list is to facilitate the discussion between users
>> of the FieldTrip  toolbox, to share experiences and to discuss new
>> ideas for MEG and EEG analysis. See also http://
>> listserv.surfnet.nl/archives/fieldtrip.html and http://www.ru.nl/
>> fcdonders/fieldtrip.
>>
>
> ----------------------------------
> The aim of this list is to facilitate the discussion between users
> of the FieldTrip  toolbox, to share experiences and to discuss new
> ideas for MEG and EEG analysis. See also http://listserv.surfnet.nl/
> archives/fieldtrip.html and http://www.ru.nl/fcdonders/fieldtrip.

----------------------------------
The aim of this list is to facilitate the discussion between users of the FieldTrip  toolbox, to share experiences and to discuss new ideas for MEG and EEG analysis. See also http://listserv.surfnet.nl/archives/fieldtrip.html and http://www.ru.nl/fcdonders/fieldtrip.