Hello,

I want to do IIR filtering in Scilab, but can't really figure out how it all
works. The Matlab code for what I want to do would look something like this:

x = some signal...
B = [1 2 3];
A = [4 5 6];
y = filter(B,A,x);

Now, that's quite simple, but how do I do the same thing in Scilab? It looks
like I should be using flts(), but I can't really figure out how to
translate my coefficient vectors A & B into whatever formatting flts()
expects.

Thanks,
Jesper

Thus spake JBK:
> Hello,
> 
> I want to do IIR filtering in Scilab, but can't really figure out how it all
> works. The Matlab code for what I want to do would look something like this:
> 
> x = some signal...
> B = [1 2 3];
> A = [4 5 6];
> y = filter(B,A,x);
> 
> Now, that's quite simple, but how do I do the same thing in Scilab? It looks
> like I should be using flts(), but I can't really figure out how to
> translate my coefficient vectors A & B into whatever formatting flts()
> expects.

This may not help you at all... but if the reason you're using Scilab is
because it's a free simulation package, you could try octave
(www.octave.org) -- it is also a free simulation package, but it is
largely code compatible with MatLab.  Virtually all the (albeit fairly
basic I imagine) simulations I've written in MatLab run without changes
in octave.  Your code above would certainly work.

I haven't tried the windows version, but there certainly is one
mentioned on the downloads page.

Might be completely useless to you if you're set on using SciLab, but I
figured it couldn't hurt!

Cheers,

Al

> This may not help you at all... but if the reason you're using Scilab is
> because it's a free simulation package, you could try octave
> (www.octave.org) -- it is also a free simulation package, but it is
> largely code compatible with MatLab.  Virtually all the (albeit fairly
> basic I imagine) simulations I've written in MatLab run without changes
> in octave.  Your code above would certainly work.
>
> I haven't tried the windows version, but there certainly is one
> mentioned on the downloads page.
>
> Might be completely useless to you if you're set on using SciLab, but I
> figured it couldn't hurt!
>
> Cheers,
>
> Al

As you've already guessed the only reason I use Scilab is that it is free. I
would definitely prefer Matlab if it hadn't been for the ridiculous price. I
know that Octave is more Matlab-compatible, but the signal-processing
toolbox in Scilab is far more extensive than the one in Octave.

Thanks anyway,
Jesper

Maybe 'frmag' will help somehow?
"JBK" <jesper.kristensen@virginNOSPAM.net> wrote in message
news:4jbajb.m5d.ln@172.16.100.88...
> > This may not help you at all... but if the reason you're using Scilab is
> > because it's a free simulation package, you could try octave
> > (www.octave.org) -- it is also a free simulation package, but it is
> > largely code compatible with MatLab.  Virtually all the (albeit fairly
> > basic I imagine) simulations I've written in MatLab run without changes
> > in octave.  Your code above would certainly work.
> >
> > I haven't tried the windows version, but there certainly is one
> > mentioned on the downloads page.
> >
> > Might be completely useless to you if you're set on using SciLab, but I
> > figured it couldn't hurt!
> >
> > Cheers,
> >
> > Al
>
> As you've already guessed the only reason I use Scilab is that it is free.
I
> would definitely prefer Matlab if it hadn't been for the ridiculous price.
I
> know that Octave is more Matlab-compatible, but the signal-processing
> toolbox in Scilab is far more extensive than the one in Octave.
>
> Thanks anyway,
> Jesper
>
>

Hi Jesper,

> I want to do IIR filtering in Scilab

I _think_ you're looking for rtitr() (I've never tried it so YMMV).


HTH,

Rick Armstrong

(reply address is bogus)

JBK wrote:
> 
> As you've already guessed the only reason I use Scilab is that it is free. I
> would definitely prefer Matlab if it hadn't been for the ridiculous price. I
> know that Octave is more Matlab-compatible, but the signal-processing
> toolbox in Scilab is far more extensive than the one in Octave.
> 

The Matlab student price is very reasonable and their policy
for qualifying as a student is too.  (At least it used to
be.)


Bob
-- 

"Things should be described as simply as possible, but no
simpler."

                                             A. Einstein

> The Matlab student price is very reasonable and their policy
> for qualifying as a student is too.  (At least it used to
> be.)

How about not being a student at all?
And can a student license be used for commercial use as well?

/Jesper

no, in the student license agreement it explicity says "non-commercial"


JBK wrote:
>>The Matlab student price is very reasonable and their policy
>>for qualifying as a student is too.  (At least it used to
>>be.)
> 
> 
> How about not being a student at all?
> And can a student license be used for commercial use as well?
> 
> /Jesper
> 
> 

there's a dedicated newsgroup
comp.soft.sys.math.scilab
you should try to ask there

good luck!

"JBK" <jesper.kristensen@virginNOSPAM.net> ha scritto nel messaggio
news:973ajb.o1e.ln@172.16.100.88...
> Hello,
>
> I want to do IIR filtering in Scilab, but can't really figure out how it
all
> works. The Matlab code for what I want to do would look something like
this:
>
> x = some signal...
> B = [1 2 3];
> A = [4 5 6];
> y = filter(B,A,x);
>
> Now, that's quite simple, but how do I do the same thing in Scilab? It
looks
> like I should be using flts(), but I can't really figure out how to
> translate my coefficient vectors A & B into whatever formatting flts()
> expects.
>
> Thanks,
> Jesper
>
>

You might also look at O-Matrix, (http://www.omatrix.com).  It's not free,
but it's quite inexpensive and provides support and other commercial
benefits.


"A. S. Budden" <newsreader.mail@NOSPAM.southparade.vispa.com> wrote in
message news:20030905141247.GA29148@zakalwe...
> Thus spake JBK:
> > Hello,
> >
> > I want to do IIR filtering in Scilab, but can't really figure out how it
all
> > works. The Matlab code for what I want to do would look something like
this:
> >
> > x = some signal...
> > B = [1 2 3];
> > A = [4 5 6];
> > y = filter(B,A,x);
> >
> > Now, that's quite simple, but how do I do the same thing in Scilab? It
looks
> > like I should be using flts(), but I can't really figure out how to
> > translate my coefficient vectors A & B into whatever formatting flts()
> > expects.
>
> This may not help you at all... but if the reason you're using Scilab is
> because it's a free simulation package, you could try octave
> (www.octave.org) -- it is also a free simulation package, but it is
> largely code compatible with MatLab.  Virtually all the (albeit fairly
> basic I imagine) simulations I've written in MatLab run without changes
> in octave.  Your code above would certainly work.
>
> I haven't tried the windows version, but there certainly is one
> mentioned on the downloads page.
>
> Might be completely useless to you if you're set on using SciLab, but I
> figured it couldn't hurt!
>
> Cheers,
>
> Al

"Harmonic Software Inc." wrote:
> 
> You might also look at O-Matrix, (http://www.omatrix.com).  It's not free,
> but it's quite inexpensive and provides support and other commercial
> benefits.
> 
A quick look at the site didn't tell me if it's {zero}-Matrix or
{oh}-Matrix. I hope for the former, and that it allows zero-based
arrays.

Jerry
-- 
Engineering is the art of making what you want from things you can get.
�����������������������������������������������������������������������

<Filtering with Scilab>

| x = some signal...
| B = [1 2 3];
| A = [4 5 6];
| y = filter(B,A,x);

z=poly(0,'z');
N=1+2/z+3/z**2
D=4+5/z+6/z**2
H=N./D
H=syslin('d',H);
y=flts(x,H);

Thanks!

I did try the dedicated Scilab newsgroup, but it appears to have very few
readers.

/Jesper

"Curl" <Mr.Bilou@microsoft.fr> wrote in message
news:3f5c2f76$0$10404$626a54ce@news.free.fr...
> <Filtering with Scilab>
>
> | x = some signal...
> | B = [1 2 3];
> | A = [4 5 6];
> | y = filter(B,A,x);
>
> z=poly(0,'z');
> N=1+2/z+3/z**2
> D=4+5/z+6/z**2
> H=N./D
> H=syslin('d',H);
> y=flts(x,H);
>
>

"Curl" <Mr.Bilou@microsoft.fr> wrote in message
news:3f5c2f76$0$10404$626a54ce@news.free.fr...
> <Filtering with Scilab>

I have a basic question about SCILAB system in here.

> z=poly(0,'z');
> N=1+2/z+3/z**2
> D=4+5/z+6/z**2
> H=N./D
> H=syslin('d',H);
> y=flts(x,H);

Curl, does this dictate that SCILAB has ability to treat
symbolic calculation sine it converts from sequence form
to poly form before filtering?  More specifically,
is the SCILAB based on symbolic mathematic engine,
like Mathematica and Maple?

In short, I asked the capability of symbolic calculation on
SCILAB and the system behind it.

-- 
Best regards,
James K. (txdiversity@hotmail.com)
- Private opinions: These are not the opinions from my affiliation.

"Jerry Avins" <jya@ieee.org> wrote in message
news:3F5B936D.AA4209B6@ieee.org...
> "Harmonic Software Inc." wrote:
> >
> > You might also look at O-Matrix, (http://www.omatrix.com).  It's not
free,
> > but it's quite inexpensive and provides support and other commercial
> > benefits.
> >
> A quick look at the site didn't tell me if it's {zero}-Matrix or
> {oh}-Matrix. I hope for the former, and that it allows zero-based
> arrays.

Yet, we have another choice that is highly compatible with
matlab, CEMtool 5.0 (http://www.cemtool.co.kr).
In addition to the language compatibility (m-mode),
it provides a great useful and integrated GUI environment,
e.g. command-liner, editor, and Web navigator.

>
> Jerry
> -- 
> Engineering is the art of making what you want from things you can get.
> �����������������������������������������������������������������������

-- 
Best regards,
James K. (txdiversity@hotmail.com)
- Private opinions: These are not the opinions from my affiliation.