Help Coherence in matlab

Can anyone help me do scaling for matlab's mscohere?

I have two time series x and Y of the same length (1x220) which I want to
know the relationships spectrally. mscohere gives me the coherence (cxy)
with a vector (1 x 146) which is lower than the length of x or y. I can't
explain it, so is there anyway I can make a better plot where I can derive
any meaning? The peaks are ok ranging from 0-1. The xaxis is not.

Regards
Brett

bluenoble4u wrote:

> Can anyone help me do scaling for matlab's mscohere?
> 
> I have two time series x and Y of the same length (1x220) which I want to
> know the relationships spectrally. mscohere gives me the coherence (cxy)
> with a vector (1 x 146) which is lower than the length of x or y. I can't
> explain it, so is there anyway I can make a better plot where I can derive
> any meaning? The peaks are ok ranging from 0-1. The xaxis is not.
> 
> Regards
> Brett
> 
> 

On Thu, 17 Sep 2009 09:08:06 -0500, "bluenoble4u"
<bluenoble4u@yahoo.com> wrote:

>Can anyone help me do scaling for matlab's mscohere?
>
>I have two time series x and Y of the same length (1x220) which I want to
>know the relationships spectrally. mscohere gives me the coherence (cxy)
>with a vector (1 x 146) which is lower than the length of x or y. I can't
>explain it, so is there anyway I can make a better plot where I can derive
>any meaning? The peaks are ok ranging from 0-1. The xaxis is not.
>
>Regards
>Brett

Hi,
   My old version of MATLAB doesn't have that built-in 
mscohere() function.  So I can't help ya' out here.
I did take a quick look at: 

http://www.mathworks.com/access/helpdesk/help/toolbox/signal/index.html?/access/helpdesk/help/toolbox/signal/mscohere.html&http://www.google.com/url?sa=t&source=web&ct=res&cd=1&url=http%3A%2F%2Fwww.mathworks.com%2Faccess%2Fhelpdesk%2Fhelp%2Ftoolbox%2Fsignal%2Fmscohere.html&ei=NMq0SvzfJ4j6sQP17NnRDA&rct=j&q=matlab+mscohere&usg=AFQjCNFaq_IEXffDyb3OMAdIPL3uzuz1Kg

It looks like the mscohere() function is fairly complicated 
and it uses Welch's Method of partitioning (breaking) input 
sequences into smaller blocks of samples on which to perform 
spectrum analysis.  Also, the web site said something 
about mscohere() "returning a one-sided coherence estimate."
I interpret that, knowing no better, that the function's output 
only covers the positive freq range.

I'd better quit here, because I really don't know enough 
about mscohere() to fill a thimble.

Good Luck,
[-Rick-]