[Mne_analysis] Requesting help for removing artifacts from Single-Channel EEG Signals

[Alexandre Gramfort]

        External Email - Use Caution        

hi,

ICA on 1 channel does not make sense. ICA applies an unmixing matrix
to data so if you have 1 channel you're doing a dot product with a float.

There exist underdetermined ICA versions but it's not implemented
in MNE and not common on EEG/MEG data.

What artifacts do you want to remove?

HTH
Alex


On Mon, Dec 10, 2018 at 2:38 PM VENKATA PHANIKRISHNA B <
b.phanikrishna at gmail.com> wrote:

>         External Email - Use Caution
>
> Hi, everyone.
> I try to do remove artifacts from free available EEG data (from
> PHYSIONET). For entire dataset, it working well by using the following code.
> from mne.preprocessing import ICA, create_ecg_epochs
> raw=mne.io.read_raw_edf(r'C:\Users\shash\OneDrive\Desktop\desktop\SC4001E0-PSG.edf',
> preload=True)
> raw_tmp = raw.copy()
> ica = mne.preprocessing.ICA(method="extended-infomax", random_state=1)
> ica.fit(raw_tmp)
>
> This procedure working only for ".edf" format file
> Same thing I try to do it for single channel data that were taken into
> another variable (let say in another format file like excel or text) I am
> getting a divide by zero error.
>
> Can anyone remove artifacts from Single-Channel EEG Signals?
> Please provide help for removing artifacts for attached EEG data file.
> Properties of ATTACHED file:
> Data: EEG amplitude
> number of channels: 1 i.e.,'EEGFpzCz'
> sampling frequency: 100HZ.
> Epochs =30seconds
> Total samples or amplitudes=30*100=300
>
>   I am looking forward to a positive response from you.
>  THANK YOU SO MUCH FOR YOUR TIME AND CONSIDERATION
>
>
>
>
>
> --
> ************************
> thanks
>
>
> *VENKATA PHANIKRISHNA B*
> 9908261261
> _______________________________________________
> Mne_analysis mailing list
> Mne_analysis at nmr.mgh.harvard.edu
> https://mail.nmr.mgh.harvard.edu/mailman/listinfo/mne_analysis
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://mail.nmr.mgh.harvard.edu/pipermail/mne_analysis/attachments/20181210/8d1396ae/attachment-0001.html