Bandpass Filtering EEG Data

Hi!

I'm trying to bandpass filter EEG data that has 64 channels. For some reason, the filter is replacing a large part of some channels with NaNs. Here is the code I'm using. Fs in this case is 240 Hz. I have attached the EEG data that I'm using. Thank you so much!

function [ppf] = pfilter(x, Fs)

%First apply a bandpass filter to get frequencies between 0.1 Hz and 40 Hz

Fn = Fs/2; %nyquist frequency
wp = [1 40]/Fn; %normalizing frequency range. 
[b,a] = butter(1,wp);%1st order butterworth filter. 

[sos,g] = tf2sos(b,a);
figure(1);
freqz(sos, 240, Fs);
ppf = x;

for i = 1:size(x,2);
    tmp = ppf(:,i); tmp(tmp==0)=NaN;
    tmp(isfinite(tmp)) = detrend(tmp(isfinite(tmp)));
    ppf(:,i) = filter(b,a,tmp);
end

end

Answer 1

[Daniel M]

Here's your problem

tmp(tmp==0)=NaN;

Don't do that.