Thread Subject: Which ML function should I use to approx smoothness of a signal?

I have a raw signal as a vector.
I have a smoothed signal derived from the raw signal with an
in-house algorithm. (It is another vector.)

I need to quantify the variance of the smoothed signal from
the raw signal. What are some good options for this in Matlab?

Also, as a second question, if I want to quantify the noise
in the raw signal (referenced to itself only), what Matlab
functions should I look at?

Thanks

"G.A.M. " <x0zero@gmail.com> wrote in message <fdrhn6$nti
$1@fred.mathworks.com>...
> I have a raw signal as a vector.
> I have a smoothed signal derived from the raw signal with an
> in-house algorithm. (It is another vector.)
>
> I need to quantify the variance of the smoothed signal from
> the raw signal. What are some good options for this in Matlab?
>
> Also, as a second question, if I want to quantify the noise
> in the raw signal (referenced to itself only), what Matlab
> functions should I look at?

Interesting. I happen to have a code that solves
exactly this problem. I'll post it on the file exchange.
I was waiting to add a few more bells and whistles,
but if there is a use for it, I'll post it now. Give it a
day to show up there.

As an example, gaussian noise added to a sine wave,
nominal variance = 0.01.

t = linspace(0,1,1000)';
x = sin(t*50) + randn(size(t))/10;

v = estimatenoise(x)
v =
      0.0096887

HTH,
John