Re: filter.oct

[Esposito Francesco]

Hi all,

i have implemented my personal filter.c based on Eigen Lib 
functions and It behaves like Matlab but I tried the 
filter function of octave but It behaves differently from 
matlab.

If you try y=filter(b,a,x,zi) both in matlab and octave 
you can appreciate this difference. Use this input 
parameters where a and b are chosen as chebyshev 
coefficients while zi(initial condition) and x(input) are 
invented just to check the function:

zi=[0.9943 -5.5456 13.7336 -19.4771 17.0401 -9.1809 2.8179 
-0.3799];
b=1.0e-04*[0.0028    0.0223    0.0780    0.1560    0.1950 
  0.1560    0.0780    0.0223    0.0028];
a=[1.0000   -6.9463   21.5646  -39.0372   45.0335 
-33.8827   16.2314   -4.5256    0.5623];
x=[0.1000 0.0100 0.2000 0.0030 0.6000 3.0000 0.5000 0.7000 
0.8800 0.5000];

Do you know the reason of this difference?

I'm also available to patch your filter.cc as you ask me 
in the previous mail.....I'm not a control-guy but an 
Embedded Engineer and I'm not so impertinent to suppose 
your filter does not work properly. Probably I did not 
understand the correct use of your filter.

Thank you all and best

Francesco


On Fri, 25 Feb 2011 08:02:20 -0600
 Jordi Gutiérrez Hermoso <address@hidden> wrote:
> On 25 February 2011 04:52, Esposito Francesco 
> <address@hidden> wrote:
> > I need to implement a filter function in C language with 
> > the SAME behaviour
> > of the Matlab one.
> >
> > I solved the problem.....
> >
> > I get the formula from 
> > http://www.mathworks.com/help/techdoc/ref/filter.html
>
> Can you give us a patch for src/DLD_FUNCTIONS/filter.cc?
>
> Thanks,
> - Jordi G. H.