I have a question related to the screenshot below. In my preprocessing
script I've gone through the following steps:
-Import raw data with *raw=mne.io.read_raw_egi(raw_file,
montage=mne.channels.read_montage(kind='GSN-HydroCel-129'),preload=True,
verbose=True)*
-filter data with *raw.filter(1,50)*
-interpolating bad electrodes: *raw.interpolate_bads()*
-rereferencing with a global avg with *mne.set_eeg_reference(raw,
ref_channels=scalp_electrodes+ext_electrodes) *
-then, *ica = ICA(n_components=90,random_state=25)*
I've also tried running it with *method=infomax*, but the problem still
persists of high-frequency noise in several IC's in every subject. This is
despite the high and low pass filtering mentioned above.
Any thoughts as to why I might still be seeing this would be greatly
appreciated.
You should try changing the `h_trans_bandwidth` to be narrower, by default
it will use 0.25 times the `h_freq` so in this case 12.5 Hz, which means
that 60 Hz is in the transition band rather than the stop-band. See (though
there is a bug in the table; the values under `h_trans_bandwidth` for 45
and 48 Hz should be 11.25 and 12, respectively):
You could also use fir_window='blackman', which will give you a longer
filter but also better stop-band attenuation. In 0.18 doing
`mne.filter.create_filter(..., verbose=True)` will give you useful
information about the filter you are using.
Eric
External Email - Use Caution
Hello MNE community,
I have a question related to the screenshot below. In my preprocessing
script I've gone through the following steps:
-Import raw data with *raw=mne.io.read_raw_egi(raw_file, montage=mne.channels.read_montage(kind='GSN-HydroCel-129'),preload=True,
verbose=True)*
-filter data with *raw.filter(1,50)*
-interpolating bad electrodes: *raw.interpolate_bads()*
-rereferencing with a global avg with *mne.set_eeg_reference(raw,
ref_channels=scalp_electrodes+ext_electrodes) *
-then, *ica = ICA(n_components=90,random_state=25)*
I've also tried running it with *method=infomax*, but the problem still
persists of high-frequency noise in several IC's in every subject. This is
despite the high and low pass filtering mentioned above.
Any thoughts as to why I might still be seeing this would be greatly
appreciated.
Thank you!
Best wishes,
Dillan Cellier
[image: Screen Shot 2019-05-20 at 12.00.07 PM.png]
_______________________________________________
Mne_analysis mailing list
Mne_analysis at nmr.mgh.harvard.edu Mne_analysis Info Page
