[Eeglablist] Transferring ICA weights when different number of channels

[Makoto Miyakoshi]

Dear Carolin,

Sorry for belated response.

> The next problem is that the interpolated channels are added at the end
of the channel list, i.e., different subjects have different channel
orders.

Are you sure? I confirmed that if you use the method I showed below, the
channel order remain to be the same.
https://sccn.ucsd.edu/wiki/Makoto's_preprocessing_pipeline#Interpolate_all_the_removed_channels

Could you please try it?

Makoto

On Sat, Feb 25, 2017 at 12:20 PM, Carolin Sievers (PSY) <
Carolin.Sievers at uea.ac.uk> wrote:

> Dear Makoto,
>
> thank you for the prompt response. I followed your advice and then
> interpolated the missing channels later. The next problem is that the
> interpolated channels are added at the end of the channel list, i.e.,
> different subjects have different channel orders.
>
> I have tried fixing this through this command:
> [~,index] = sortrows([EEG.chanlocs.urchan].'); EEG.chanlocs =
> EEG.chanlocs(index); clear index
> However, this only changes the labels and locations but not the actual
> EEG.data structure. I have come across some posts related to this issue in
> the mailing list archives but haven’t yet found a solution. I image the
> reshape command may be of use but due to the sheer size of EEG.data I find
> it difficult to tackle this problem.
>
> Do you happen to have any solution for this? My final idea would be to
> calculate ERPs and then go through channel operations but ideally, I would
> like to change channel order to the EEG data structure and not just the ERP.
>
> All the best,
> Carolin
>
> On 10 Feb 2017, at 03:10, Makoto Miyakoshi <mmiyakoshi at ucsd.edu> wrote:
>
> Dear Carolin,
>
> > How do I account for that difference in channels when transferring the
> ICA weights matrix? Is there any way of keeping the excluded channels
> separate and add them back in after removal of ICs?
>
> It's better to let EEGLAB take care of these consistency.
> I recommend that reject the same channels between sender and receiver to
> make sure that they have the identical channel structure, then copy the
> weight matrix.
>
> Makoto
>
>
> On Tue, Feb 7, 2017 at 12:53 PM, Carolin Sievers (PSY) <
> Carolin.Sievers at uea.ac.uk> wrote:
>
>> Dear EEGLAB users,
>>
>>
>>
>> I have a question regarding transferring ICA weights from one dataset to
>> another (e.g., high-pass filter at 1Hz for ICA, then apply weights matrix
>> to same data but filtered at 0.05Hz):
>>
>>
>>
>> For better ICA results, I sometimes exclude some noisy channels. How do I
>> account for that difference in channels when transferring the ICA weights
>> matrix? Is there any way of keeping the excluded channels separate and add
>> them back in after removal of ICs?
>>
>>
>>
>> Best wishes,
>>
>> Carolin
>>
>>
>> _________________
>>
>> *Carolin Sievers* (PSY)
>> PhD Student
>> School of Psychology, LSB 0.111
>>
>> University of East Anglia – Norwich Research Park
>>
>> Norwich, NR4 7TJ, United Kingdom
>>
>>
>>
>> *Carolin.Sievers at uea.ac.uk <Carolin.Sievers at uea.ac.uk>*
>>
>> https://www.uea.ac.uk/psychology/people/profile/carolin-sievers
>> *https://twitter.com/carolinsmemory <https://twitter.com/carolinsmemory>*
>>
>>
>>
>>
>>
>>
>>
>> <image001.png>
>>
>>
>>
>> *UK Top 15* (Times and Sunday Times Good University Guide 2017)
>>
>> *Top 5 for student satisfaction* (National Student Survey 2005-2016)
>>
>> *World Top 1%* (Times Higher Education World Rankings 2016-17)
>>
>> *World Top 100 for research excellence* (Leiden Ranking 2016)
>>
>> This email is confidential and may be privileged. If you are not the
>> intended recipient please accept my apologies;
>>
>> please do not disclose, copy or distribute information in this email or
>> take any action in reliance on its contents:
>>
>> to do so is strictly prohibited and may be unlawful. Please inform me
>> that this message has gone astray before deleting it.
>>
>> Thank you for your co-operation.
>>
>>
>>
>> _______________________________________________
>> Eeglablist page: http://sccn.ucsd.edu/eeglab/eeglabmail.html
>> To unsubscribe, send an empty email to eeglablist-unsubscribe at sccn.uc
>> sd.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
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://sccn.ucsd.edu/pipermail/eeglablist/attachments/20170510/ba12bf52/attachment.html>