filtfilt() function with Matlab Coder

Hello,

I am trying to get the Matlab Coder to convert my matlab code to C code.

And there is a problem with the filtfilt() function I do not understand.

The error that the Coder gives me is the following: "The numerator coefficients B and denominator coefficients A must be constant."

I get the coefficients from a butterworth filter, so basically I do the following:

[b, a] = butter(n,Wn,'low');

euler(:,3) = filtfilt(b,a,euler(:,3));

I even tried just grabbing the values and define a and b before calling the function like this:

a = [1 -3.37801139459687 4.75177537470702 -3.43971331083559 1.27399988148292 -0.192388596001173];
b = [0.000489436086134679 0.00244718043067339 0.00489436086134679 0.00489436086134679 0.00244718043067339 0.000489436086134679];
euler(:,3) = filtfilt(b,a,euler(:,3));

But it still gives me the same error. I really don't know how to make this work, can somebody please help me?

[Mike Hosea]

I don't understand it, either. This works fine for me in 16b:

function e = tff(e)
n = 5;
Wn = .3;
[b, a] = butter(n,Wn,'low');
e = filtfilt(b,a,e);

The variables n and Wn are constant, right? Are you re-using the variables a and b for other purposes? Used coder.varsize on these variable names?

Answer 1

[feng cai]

if the code like this:

b and a is defined outside the function bandpass

function filt_data = bandpass(b,a,data)

filt_data = filtfilt(b,a,data);

end

Error: The numerator coefficients B and denominator coefficients A must be constant.

----------------------------------------------------------------------------------

if the code like this:

function filt_data = bandpass(data)

a = [1,-2.86834067990259,3.20377999626534,-1.70427191678698,0.378619299408443];

b= [0.0790937181169893,0,-0.158187436233979,0,0.0790937181169893];

filt_data = filtfilt(b,a,data);

end

Source Code generation succeeded.