COMPGROUPS.NET | Search | Post Question | Groups | Stream | About | Register

### mutual information #2 420262

• Email
• Follow

```hi all,

i have a doubt regarding mutual information. this file computes the probability of mutual information occuring in two signals. but why sud we compute the sum of probabilities at the last.I=sum(I) ... shud we plot this probability of various signal comparisons with time delay to get optimum lag... pls reply...

A=double(A);
B=double(B);

na = hist(A(:),2000);
na = na/sum(na);

nb = hist(B(:),2000);
nb = nb/sum(nb);

n2 = hist2(A,B,2000);
n2 = n2/sum(n2(:));

I=minf(n2,na'*nb);
I=sum(I);

thanks
kirthi
```
 0

See related articles to this posting

```"kiruthiga" wrote in message <ii5s9j\$an\$1@fred.mathworks.com>...
> hi all,
>
> i have a doubt regarding mutual information. this file computes the probability of mutual information occuring in two signals. but why sud we compute the sum of probabilities at the last.I=sum(I) ... shud we plot this probability of various signal comparisons with time delay to get optimum lag... pls reply...
>
> A=double(A);
> B=double(B);
>
> na = hist(A(:),2000);
> na = na/sum(na);
>
> nb = hist(B(:),2000);
> nb = nb/sum(nb);
>
> n2 = hist2(A,B,2000);
> n2 = n2/sum(n2(:));
>
> I=minf(n2,na'*nb);
> I=sum(I);
>
> thanks
> kirthi

Hi, minf.m is not a MATHWORKS function. Is this a part of some toolbox you have downloaded?

Wayne
```
 0

```hi wayne

s... minf is another matlab program... from mutual information toolbox... is that we have to plot various probablities of MI vs optimum lag...?

kirthi
```
 0