Low Pass Filter returning NaN

New user, here...

I'm trying to use a LPF on an array with data from a sensor. The sampling rate of the data is either 375/sec or 500/sec. The samples are between 1000/-1000.

I was successfully using the following script on one set of data:

%Low Pass Filter
f=500;
fNormLow = 2/(f/2);
[b1,a1] = butter(2,fNormLow,'low');
lowPass = filtfilt(b1,a1,rawData(:,1));
figure(3);
plot(lowPass(:,1),'-m')
hold on;

However, the same script doesn't work on a different set of data. lowPass(:,1) is entirely filled with NaN.

Can anyone lend some insight as to what I'm doing wrong? I'm thinking I need to try a FIR filter instead of an IIR, but I'm not sure why it works for the first set of data and not for the second. If the answer is a different filter, I'm hoping for a suggestion of which one to use. I'm looking for a very smooth sine wave to be made out of data with a periodic spike - pretty simple. Maybe someone could post an example?

Thanks in advance!

[Honglei Chen]

You may want to check your raw data and see if it has things like NaN or Inf in there.

[Ethan]

Thanks for the quick response!

There is one blank row in the txt file (when imported in to excel) -- however, there is more than one blank row in the data file that works. I used the "isnumber" function in excel to check if there was anything other than a number, and that blank row was the only hit.