why filtering gives nan values.

Am applying low pass filter to my signal data using

 [b11,a11] = butter(5,4/24);
  MM1= filtfilt(b11,a11,MM1);

but the result comes out nan. I don't have any nans in my MM1 data . Any help?

Answer 1

[Youssef Khmou]

hi,

I think the input has a NaN, look at this example :

 r=rand(100,1);
 [b11,a11] = butter(5,4/24);
 MM1= filtfilt(b11,a11,r); % Normal Filtering

r(end/2)=NaN;    %  FOR example a middle value to be NaN
MM2= filtfilt(b11,a11,r);

MM2 is vector of NaNs, only one NaN value in r produces NaNs in the output .

try :

 isnan(MM1)

[Ede gerlderlands]

Thanks, I checked that there is no nan in the input.

[Youssef Khmou]

hi, well that is not usual, ok can you process only the Half of the MM1 vector ans see ? can you post the vector MM1?

[Honglei Chen]

@Ede, do you have things like inf in your data?

[Ede gerlderlands]

Thanks for your support