[Eeglablist] How to get same time points of different epoch in the function newtimef?

[Makoto Miyakoshi]

Dear anonymous Chinese friend,

> But if the value cannot be divisible by “timesout”, how newtimef
determined the times point output?

I don't know how it behaves under that situation.

Isn't it possible to use the same data length? I think that's the most
straightforward solution for you. You don't need to worry about the overlap
of the data, just epoch everything into from -1 to 2 sec epoch and use
parameters 'freqs', [3 100] and cycle [3 0.85].

Makoto

On Wed, Apr 1, 2015 at 8:22 PM, <methodlearning at sina.cn> wrote:

> Dear Makoto,
>
> Thank you !
>
> When the value of “End Time-(window length)/2-Start Time+(window
> length)/2” can be divisible by “timesout”, the calculation method of time
> points output sounds clearly.
>
> But if the value cannot be divisible by “timesout”, how newtimef
> determined the times point output?
>
> e.g.
>
> 1.      250ms window length
>
> 2.      from -1000ms to 2000ms epoch length
>
> 3.      200 timesout
>
> The calculation method is not  “round”/’Ceil” or “floor” according to my
> own experience. When the value can’t be divisible, the time interval of
> adjacent time points seems not to be fixed.
>
>                                                                      I.P.
> ----- 原始邮件 -----
> 发件人：Makoto Miyakoshi <mmiyakoshi at ucsd.edu>
> 收件人：methodlearning at sina.cn
> 抄送人：eeglablist <eeglablist at sccn.ucsd.edu>
> 主题：Re: Re: [Eeglablist] How to get same time points of different epoch in
> the function newtimef?
> 日期：2015年04月02日 05点50分
>
> Dear anonymous friend,
>
> The superficial temporal resolution is determined in this way.
>
> If
>
>    1. 1000ms window length (i.e. half of its length will be removed from
>    the both ends of the epoch length)
>    2. from -1000ms to 2000ms epoch length
>    3. 200 timesout
>
> then you'll see -500ms to 1500ms divided by 200 == 10 ms time resolution
> in the output plot. If you want to have the same output, you need to feed
> the same length of the data, which should be possible.
>
> See slide 21 of this pdf
>
> http://sccn.ucsd.edu/mediawiki/images/1/19/C2_A3_Time-frequencyDecAndAdvancedICAPracticum.pdf
>
> Makoto
>
>
>
> On Tue, Mar 31, 2015 at 2:40 AM, <methodlearning at sina.cn> wrote:
>
> Hi Makoto,
>
> Thanks a lot for your reply!
>
> If I use the function newtimef to decompose epochs of different conditions
> and give input parameter "timesout" the same value, the time points output
> after 0 ms will be very different across conditions because the time spans
> were different .For example, if I decompose different epochs using the
> same function newtimef(EEG.data(elec,:,:)EEG.pnts,[EEG.xmin
> EEG.xmax]*1000,500,’cycles’,[2
> 0.5],'baseline',BaselineVector,'timesout',250,'nfreqs',93,'winsize',250),the
> time points output after 0 ms for condition A will be :[6 14 20 26 34 40 46
> 52 60 66 72 80 86 92...];
>
> and condition B :[4 12 22 30 38 46 54 62 72 80 ...];
>
> and condition C:[4 10 16 22 28 34 40 46...].
>
> However, I need to compare ersp value across conditions on the same time
> points using ANOVA after decomposing. The decomposing results of condition
> A mentioned above included the ersp data at the time point"6", and results
> of condition B and C didn't include the time point"6",so we can't do ANOAV.
> And I am not interested in the ersp value before 0 ms .
>
>  So I need to know how to get ersp value on one-one corresponding time
> points after 0 ms across conditions of different epoch span.Maybe I need
> to know newtimef algorithm of determining time points output ,to adjust the
> input parameter "timesout", to get ersp value on one-one corresponding time
> points across conditions(after 0 ms) .
>
> I am not sure if I made myself clear now?
>
> Thank you again for your help!
>
>
>
>   I.P.
> ----- 原始邮件 -----
> 发件人：Makoto Miyakoshi <mmiyakoshi at ucsd.edu>
> 收件人：methodlearning at sina.cn
> 抄送人：eeglablist <eeglablist at sccn.ucsd.edu>
> 主题：Re: [Eeglablist] How to get same time points of different epoch in the
> function newtimef?
> 日期：2015年03月31日 03点40分
>
> Dear I.P.,
>
> I need more info (data) to answer to your question. What do you mean by
> 'so different'?
>
> Makoto
>
> M
>
> On Sat, Mar 28, 2015 at 10:34 PM, <methodlearning at sina.cn> wrote:
>
> Hi all,
>
>       In the time-frequency analysis,the epoch methods of different
> conditions were different:
>
> The epoch of condition A was [-1400 800]ms,and the baseline was [-1300
> -1000]ms;
>
> The epoch of condition B was [-1832 800]ms ,and the baseline was[-1732
> -1432]ms;
>
> The epoch of condition C was [-1220 800]ms ,and the baseline was[-1120
> -820]ms;
>
>    I am not interested in the  decomposition results before 0 ms,but I
> need to compare above conditions on the time points after 0 ms.
>
> I did this using the function newtimef,but the output time points of
> conditions were so different that I can't compare these conditions .How can
> I get same time points after 0 ms of different epoch?
>
> Thank you ! I am very appreciated your answers!
>
>
>                                                     I.P.
>
> _______________________________________________
> Eeglablist page: http://sccn.ucsd.edu/eeglab/eeglabmail.html
> To unsubscribe, send an empty email to
> 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
>
>
>
>
> --
> 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/20150402/9da7bce9/attachment.html>