[Eeglablist] Time- frequency analysis

Wed Aug 1 11:43:55 PDT 2018

Dear Iryna,

I'm disappearing from the list until November, so this is my last response.
Another lab member will answer the questions in other months.
First of all, I noticed several issues in your parameter settings. I
recommend you add the following optional inputs.

'cycles', [3 8], 'freqs', [1 50], 'nfreqs', 50, 'timesout', 10954

The parameter 'timesout' was calculated as (418384+647616+32768-3342)/100,
which means 'Give me data point for every 100 ms given that the sliding
window length is 3342 ms. In the current parameter setting, you have data
only this much

> Generating 200 time points (418384.0 to 647616.0 ms)

About the main baseline problem, I tried to replicate your problem. Look at
my output log below, particularly the highlighted part.
Because my continuous data (naturally) start from latency 0, the output log
says 'from 0 ms before to XXX ms after the timelocking event.'
But your log says 'from 402000 ms before to  664000 ms after the
timelocking event.' I think this is related to the problem.
Does your continuous data start from latency zero? Did you try to 'epoch'
that continuous data? I recommend you prepare the continuous data so that
it starts from latency 0. Also, in EEGLAB, continuous data cannot be
handled as single-epoched data (I don't know why...)

Makoto

> figure; pop_newtimef( EEG, 0, 1, [0  1684998], [3  8] , 'topovec',
EEG.icawinv(:,1),...
    'elocs', EEG.chanlocs, 'chaninfo', EEG.chaninfo, 'caption', ['IC 1'],
'baseline',[0 3000],...
    'nfreqs', 50, 'freqs', [1 50], 'plotphase', 'off', 'padratio', 1);
Warning: 'nfreqs' input overwrite 'padratio'
Computing Event-Related Spectral Perturbation (ERSP) and
  Inter-Trial Phase Coherence (ITC) images based on 1 trials
  of 842499 frames sampled at 500 Hz.
Each trial contains samples from 0 ms before to
  1684998 ms after the timelocking event.
  Image frequency direction: normal
Using 3 cycles at lowest frequency to 8 at highest.
Generating 200 time points (1671.0 to 1683327.0 ms)
Finding closest points for time variable
Time values for time/freq decomposition is not perfectly uniformly
distributed
The window size used is 1671 samples (3342 ms) wide.
Estimating 50 linear-spaced frequencies from 1.0 Hz to 50.0 Hz.
Processing time point (of 200): 10 20 30 40 50 60 70 80 90 100 110 120
 130 140 150 160 170 180 190 200
Computing the mean baseline spectrum
Warning: When plotting pvalues in totoplot, use option 'conv' to minimize
extrapolation effects
Note: Add output variables to command line call in history to
      retrieve results and use the tftopo function to replot them

On Wed, Aug 1, 2018 at 9:31 AM Gumenchuk, Iryna <Iryna_Gumenchuk at nymc.edu>
wrote:

> Permutation analysis will use data in (pre-0) baseline subwindows only.
> Computing Event-Related Spectral Perturbation (ERSP) and
>   Inter-Trial Phase Coherence (ITC) images based on 1 trials
>   of 131001 frames sampled at 500 Hz.
> Each trial contains samples from 402000 ms before to
>   664000 ms after the timelocking event.
> Only significant values (permutation statistics p<0.01) will be colored;
>   non-significant values will be plotted in green
>   Image frequency direction: normal
> Adjust min freq. to 6.01 Hz to match FFT output frequencies
> Adjust max freq. to 49.99 Hz to match FFT output frequencies
> Using hanning FFT tapering
> Generating 200 time points (418384.0 to 647616.0 ms)
> Finding closest points for time variable
> Time values for time/freq decomposition is not perfectly uniformly
> distributed
> The window size used is 16384 samples (32768 ms) wide.
> Estimating 1442 linear-spaced frequencies from 6.0 Hz to 50.0 Hz.
> finding closest frequencies: 6750 freqs removed
>
> Then when we use alternate settings to specify window length based on our
> sampling rate (which was 500Hz), we get:
>
> same error (see attachment )
>
>
> ________________________________
> From: Makoto Miyakoshi [mmiyakoshi at ucsd.edu]
> Sent: Tuesday, July 31, 2018 3:29 PM
> To: Gumenchuk, Iryna; EEGLAB List
> Subject: Re: [Eeglablist] Time- frequency analysis
>
> Dear Iryna,
>
> Set [250] or [251] instead of [0] in the 'sub epoch...'
> Also, what is the exact sliding window width? Please copy and paste all
> the the output report that shows up in the command line. It shows the
> window width. [250] was just an example.
>
> Makoto
>
> On Tue, Jul 31, 2018 at 11:52 AM Gumenchuk, Iryna <
> Iryna_Gumenchuk at nymc.edu<mailto:Iryna_Gumenchuk at nymc.edu>> wrote:
> Hi Makoto,
>
> Still doesn't work. I attached two images: one, where I put all setting,
> and second - the error message. May be it will give you some idea.
> Thank you!
>
> -I
> ________________________________
> From: Makoto Miyakoshi [mmiyakoshi at ucsd.edu<mailto:mmiyakoshi at ucsd.edu>]
> Sent: Tuesday, July 31, 2018 2:24 PM
> To: Gumenchuk, Iryna; EEGLAB List
> Subject: Re: [Eeglablist] Time- frequency analysis
>
> Dear Iryna,
>
> > When I do disable baseline subtraction, then it works fine
>
> Ok so the issue is in using the baseline option.
>
> > e.g., [0 30000], I get the foll0wing error message:
>
> Let's try this together.
> If your sliding window length is 500 datapoint long, try
>
> [500  30000]
>
> It should work. It may also work
>
> [250 30000]
>
> or
>
> [251 30000]
>
> I think at least one of the above parameter choices should work, if I am
> guessing the cause of the problem correctly. Basically, your time-freq data
> will start from [250] or [251] which is the half length of your sliding
> window and no data before that.
>
> Makoto
>
>
> On Mon, Jul 30, 2018 at 7:36 AM Gumenchuk, Iryna <Iryna_Gumenchuk at nymc.edu
> <mailto:Iryna_Gumenchuk at nymc.edu><mailto:Iryna_Gumenchuk at nymc.edu<mailto:
> Iryna_Gumenchuk at nymc.edu>>> wrote:
> Thank you for your response!
> Yes, this large value is for continuous data, and I want to subtract the
> baseline and then apply time - frequency analysis on continuous data. When
> I do disable baseline subtraction, then it works fine, but I need
> specifically to subtract the baseline.
> Best,
> Iryna
> ________________________________
> From: Makoto Miyakoshi [mmiyakoshi at ucsd.edu<mailto:mmiyakoshi at ucsd.edu
> ><mailto:mmiyakoshi at ucsd.edu<mailto:mmiyakoshi at ucsd.edu>>]
> Sent: Friday, July 27, 2018 9:28 PM
> To: Gumenchuk, Iryna
> Cc: EEGLAB List
> Subject: Re: [Eeglablist] Time- frequency analysis
>
> Dear Iryna,
>
> > When I use the eeglab, [0 1095000] is the value that is selected by
> default in the time-frequency dialog (pop_newtimef()).
>
> Hmm that seems a very large value, whatever it is. What is that value? My
> guess is you tried to apply time-frequency analysis on continuous data, but
> is that what you wanted to do?
>
> Also, try 'baseline', NaN to disable baseline subtraction. What do you see?
>
> Makoto
>
> On Mon, Jul 23, 2018 at 1:53 PM Gumenchuk, Iryna <Iryna_Gumenchuk at nymc.edu
> <mailto:Iryna_Gumenchuk at nymc.edu><mailto:Iryna_Gumenchuk at nymc.edu<mailto:
> Iryna_Gumenchuk at nymc.edu>><mailto:Iryna_Gumenchuk at nymc.edu<mailto:
> Iryna_Gumenchuk at nymc.edu><mailto:Iryna_Gumenchuk at nymc.edu<mailto:
> Iryna_Gumenchuk at nymc.edu>>>> wrote:
> When I use the eeglab, [0 1095000] is the value that is selected by
> default in the time-frequency dialog (pop_newtimef()). In the
> case of my continuous data, when I change select values for my baseline,
> e.g., [0 30000], I get the foll0wing error message:
> There are no sample points found in the default baseline. This may happen
> even though data time limits overlap with the baseline period (because of
> the time-freq. window width). Either disable the baseline, change the
> baseline limits.
> Please advice!
> _______________________________________________
> Eeglablist page: http://sccn.ucsd.edu/eeglab/eeglabmail.html<
> https://urldefense.proofpoint.com/v2/url?u=http-3A__sccn.ucsd.edu_eeglab_eeglabmail.html&d=DwMFaQ&c=A51OX6aSaU1ywwq_3bUC2Q&r=859eSOArK9bci42aKNCO5VGxybuCzdUW89Ib8dm8fdA&m=KNgnsr279sxq-MOl5wWMqFMzYqfdSC1TKB0LhaLi9RY&s=1FCXU7zik88Y5aMyQ7zbLMHiX2kc2Wu7-o3wAT6zQDQ&e=
> ><
> https://urldefense.proofpoint.com/v2/url?u=http-3A__sccn.ucsd.edu_eeglab_eeglabmail.html&d=DwMFaQ&c=A51OX6aSaU1ywwq_3bUC2Q&r=859eSOArK9bci42aKNCO5VGxybuCzdUW89Ib8dm8fdA&m=OCCejsqHE6dULxDTtP2dpX4fCCwyucVFYvKyNkvCxvQ&s=7oyEWzpf0BhRiUfl4WaqdUcsqGU909sPTGn9IziBGZk&e=
> ><
> https://urldefense.proofpoint.com/v2/url?u=http-3A__sccn.ucsd.edu_eeglab_eeglabmail.html&d=DwMFaQ&c=A51OX6aSaU1ywwq_3bUC2Q&r=859eSOArK9bci42aKNCO5VGxybuCzdUW89Ib8dm8fdA&m=onrp3-fYL5YNx-dBls873Uw0NLFSwVXz68zNGKtYsDE&s=iqWSNlNEEWgZDnNxZ-IxHXGTgWiA_auNBP4T__2DD0g&e=
> >
> To unsubscribe, send an empty email to
> eeglablist-unsubscribe at sccn.ucsd.edu<mailto:
> eeglablist-unsubscribe at sccn.ucsd.edu><mailto:
> eeglablist-unsubscribe at sccn.ucsd.edu<mailto:
> eeglablist-unsubscribe at sccn.ucsd.edu>><mailto:
> eeglablist-unsubscribe at sccn.ucsd.edu<mailto:
> eeglablist-unsubscribe at sccn.ucsd.edu><mailto:
> eeglablist-unsubscribe at sccn.ucsd.edu<mailto:
> eeglablist-unsubscribe at sccn.ucsd.edu>>>
> For digest mode, send an email with the subject "set digest mime" to
> eeglablist-request at sccn.ucsd.edu<mailto:eeglablist-request at sccn.ucsd.edu
> ><mailto:eeglablist-request at sccn.ucsd.edu<mailto:
> eeglablist-request at sccn.ucsd.edu>><mailto:eeglablist-request at sccn.ucsd.edu
> <mailto:eeglablist-request at sccn.ucsd.edu><mailto:
> eeglablist-request at sccn.ucsd.edu<mailto:eeglablist-request at sccn.ucsd.edu
> >>>
>
>
> --
> Makoto Miyakoshi
> Swartz Center for Computational Neuroscience
> Institute for Neural Computation, University of California San Diego
>
>
> --
> Makoto Miyakoshi
> Swartz Center for Computational Neuroscience
> Institute for Neural Computation, University of California San Diego
>
>
> --
> Makoto Miyakoshi
> Swartz Center for Computational Neuroscience
> Institute for Neural Computation, University of California San Diego
>

-- 
Makoto Miyakoshi
Swartz Center for Computational Neuroscience
Institute for Neural Computation, University of California San Diego
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://sccn.ucsd.edu/pipermail/eeglablist/attachments/20180801/159ab403/attachment.html>