I have a datafile contains data sets (size 200* 15) .how can I apply on all the data a lowpass butterworth filter of 2-order with cutoff fre 10 HZ

How can I call filter function like [b,a] = butter(n,Wn) or [b,a] = butter(n,Wn,ftype) for whole data file and collect them in another for further use.

Answer 1

[David J. Mack]

Hey Rahul,

use FILTER with the coefficients created with BUTTER, e.g.:

%Assuming X is your 200x15 data matrix and fSInHz is the sampling rate of the data.
filtFCInHz = 10; %Cut-off frequency.
filtOrder  = 2;  %Order.

[b,a] = butter(filtOrder,filtFCInHz/(fSInHz/2)); %Normalize against Nyquist frequency.
Y = filter(b,a,X);                               %Y is the filtered 200x15 output.

If you want to apply zero-phase filtering (which is recommended for time series), use FILTFILT instead of FILTER.

Greetings, David

[Rahul Pandey]

Thanks a lot David .

[David J. Mack]

Happy to help!

[Rahul Pandey]

what is the best way to find the sampling rate of data for using with filters like lowpass butterworth filter(two pass - forward and backward Zero-phase digital filtering) of 2-order with cutoff freq 10 HZ that u answered.

[David J. Mack]

Do you have a time vector for your data? If so, use:

fSInHz = mode(diff(t,[],1),1);

Assuming t is a 200xk column-oriented array with the time for each sample in seconds. If k==1 (uniform sampling rate) fSInHz is a scalar, if k==15 (different sampling rates for each column in X) it is a 1x15 row vector.