MATLAB Central Newsreader - entropy to signal

entropy to signal

george — Sun, 07 Jun 2009 21:23:01 -0400

How can I compute entropy?not to an image but to a signal?

Re: entropy to signal

Sprinceana — Sun, 07 Jun 2009 21:39:01 -0400

"george " <qwerty6542@yahoo.gr> wrote in message <h0hb3l$k76$1@fred.mathworks.com>...
> How can I compute entropy?not to an image but to a signal?

Here is a link that can help you:

http://www.cs.rug.nl/~rudy/matlab/doc/entropy.html

Re: entropy to signal

Aino — Thu, 03 Jun 2010 00:04:05 -0400

Hello,

I am trying to accomplish the same as george. I have found several codes that all claim that they calculate signal entropy, but interestingly I haven't been able to calculate the same or even closely the same entropy with the codes.

That previous link to Matlab's signal entropy won't help me, but the m-file would. Unfortunately my Matlab has also a function called entropy, but it is for a greyscale picture:

http://www.mathworks.com/access/helpdesk/help/toolbox/images/entropy.html

Could anyone please tell me where I could get that Matlab's signal entropy m-file that Sprinceana pointed out?

Thank you,
Aino

"Sprinceana " <mihaispr@yahoo.com> wrote in message <h0hc1l$li7$1@fred.mathworks.com>...
> "george " <qwerty6542@yahoo.gr> wrote in message <h0hb3l$k76$1@fred.mathworks.com>...
> > How can I compute entropy?not to an image but to a signal?
>
> Here is a link that can help you:
>
> http://www.cs.rug.nl/~rudy/matlab/doc/entropy.html

Re: entropy to signal

Walter Roberson — Thu, 03 Jun 2010 00:16:46 -0400

Aino wrote:

> Could anyone please tell me where I could get that Matlab's signal
> entropy m-file that Sprinceana pointed out?

The link to the source was at the bottom of that page:

http://www.cs.rug.nl/~rudy/matlab/source/entropy.m

Note that that is not a Mathworks file: it was written by Rudy Moddemeijer
http://www.cs.rug.nl/~rudy/matlab/

Re: entropy to signal

Aino — Thu, 03 Jun 2010 00:24:21 -0400

Wow,

I must have been blind.. Thank you. :)

-Aino

Walter Roberson <roberson@hushmail.com> wrote in message <PvCNn.39391$Ak3.37102@newsfe16.iad>...
> Aino wrote:
>
> > Could anyone please tell me where I could get that Matlab's signal
> > entropy m-file that Sprinceana pointed out?
>
> The link to the source was at the bottom of that page:
>
> http://www.cs.rug.nl/~rudy/matlab/source/entropy.m
>
> Note that that is not a Mathworks file: it was written by Rudy Moddemeijer
> http://www.cs.rug.nl/~rudy/matlab/

Re: entropy to signal

Aino — Thu, 03 Jun 2010 00:51:21 -0400

So, just to be safe, it is reasonable that the following signal has an entropy of 2.1472?:

y=1*sin(2*[1:100])+2*sin(0.5*[1:100]).^2+3*sin(0.1*[1:100]);

This is what I got using that code. I feel that entropy seems a rather random measure..

-Aino

"Aino" <aino.tietavainen@removeThis.helsinki.fi> wrote in message <hu6snl$c8q$1@fred.mathworks.com>...
> Wow,
>
> I must have been blind.. Thank you. :)
>
> -Aino
>
> Walter Roberson <roberson@hushmail.com> wrote in message <PvCNn.39391$Ak3.37102@newsfe16.iad>...
> > Aino wrote:
> >
> > > Could anyone please tell me where I could get that Matlab's signal
> > > entropy m-file that Sprinceana pointed out?
> >
> > The link to the source was at the bottom of that page:
> >
> > http://www.cs.rug.nl/~rudy/matlab/source/entropy.m
> >
> > Note that that is not a Mathworks file: it was written by Rudy Moddemeijer
> > http://www.cs.rug.nl/~rudy/matlab/

Re: entropy to signal

Walter Roberson — Thu, 03 Jun 2010 04:02:06 -0400

Aino wrote:
> So, just to be safe, it is reasonable that the following signal has an
> entropy of 2.1472?:
>
> y=1*sin(2*[1:100])+2*sin(0.5*[1:100]).^2+3*sin(0.1*[1:100]);
>
> This is what I got using that code. I feel that entropy seems a rather
> random measure..

Sorry, no idea; it would take me a bit of time to work that out, and I'd
probably be using a different entropy measure.

Re: entropy to signal

Steven Lord — Fri, 04 Jun 2010 02:32:53 -0400

"Aino" <aino.tietavainen@removeThis.helsinki.fi> wrote in message
news:hu6ua9$qb3$1@fred.mathworks.com...
> So, just to be safe, it is reasonable that the following signal has an
> entropy of 2.1472?:
>
> y=1*sin(2*[1:100])+2*sin(0.5*[1:100]).^2+3*sin(0.1*[1:100]);
>
> This is what I got using that code. I feel that entropy seems a rather
> random measure..

You might want to try working out the examples given on the documentation
page for that function on your own, to see if your results match those of
the function.

http://www.cs.rug.nl/~rudy/matlab/doc/entropy.html

The references on that page may help you understand what the code is doing
as well.

--
Steve Lord
slord@mathworks.com
comp.soft-sys.matlab (CSSM) FAQ: http://matlabwiki.mathworks.com/MATLAB_FAQ
To contact Technical Support use the Contact Us link on
http://www.mathworks.com