How to implement a generic 'filter' function using direct form I?

Asked by vinicius lanziotti

vinicius lanziotti (view profile)

on 14 Jun 2019
Latest activity Answered by Walter Roberson

Walter Roberson (view profile)

on 14 Jun 2019
Accepted Answer by Walter Roberson

Walter Roberson (view profile)

I need a code for the filter function in direct form I, like as I implemented a filter function in direct form II transposed below:
function[y] = filtro(a,b,x)
% Generic filter function in direct form II transposed
N = size(b,2) - 1;% filter order
y = zeros(size(x)); % Initializes y with zeros
%% FOR THE CASE FIR
if not(size(a,2) == size(b,2))
a(1) = 1;
a = zeros(size(b));
end
%% Recursive implementation of equations to differences
aux = zeros(1,N);
for r = 1 :size(x,2)
y(r) = b(1)*x(r) + aux(1);
for n = 2 : N
aux(n-1) = b(n)*x(r) + aux(n);
aux(n-1) = aux(n-1)- a(n)*y(r);
end
aux(N) = b(N+1)*x(r) - a(N+1)*y(r);
end
end % END OF FUNCTION
Thanks!

Walter Roberson

Walter Roberson (view profile)

on 14 Jun 2019
What is the difference between this question and the other one you asked about these filters?
vinicius lanziotti

vinicius lanziotti (view profile)

on 14 Jun 2019
The difference is that if I use the command "edit(which('dfilt.df1'))" the function has only 2 inputs (num, den), and use others parameters already defined in Matlab, like (dfilt).
% Using edit(which('dfilt.df1'))
function Hd = df1(num,den)
Hd = dfilt.df1;
Hd.ncoeffs = [1 1];
Hd.FilterStructure = 'Direct-Form I';
Hd.tapIndex = [0 0];
if nargin>=1
Hd.Numerator = num;
end
if nargin>=2
Hd.Denominator = den;
end
But I need of 3 inputs, like the filter function. But this function was made using direct-form II transposed, and I want the direct-form I from function "filter". And no one parameters already defined in Matlab.

Products

Answer by Walter Roberson

on 14 Jun 2019