Some Matlab Help

ixie99 · #1 $Old$ November 2nd, 2009, 03:46 PM

Hi all. I am new here. I though this would be a good place to ask for some help with my Matlab assignment, because i am completely lost.

I need to explore the detection of a signal in the presence of noise, the signal being the first 90 degrees of a sine wave. I need to analyze how good a fit of a polynomial with order and degree as a function increasing noise.

Using an appropriate range of signal to noise ratios, i need to construct the signal for 10 different signal to noise ratios ranging from 10:1 to 0.1:1. I need to generate polynomial fits to the data ranging from order 1 to order 4.

The last part is to calculate the residuals for each combination and demonstrate when a higher order fit is better than a linear fit.

I really need help on this. Any amount will do.
Thanks

elbarto · #2 $Old$ November 2nd, 2009, 04:31 PM

Try polyfit and polyeval.

Code:

EDU>> x = sort(rand(1,5))

x =

    0.4456    0.4898    0.6463    0.7094    0.7547

EDU>> y = rand(1,5)

y =

    0.2760    0.6797    0.6551    0.1626    0.1190

EDU>> p = polyfit(x,y,4)

p =

  1.0e+003 *

    1.0569   -2.3779    1.9403   -0.6777    0.0857

EDU>> yp=polyval(p,x)

yp =

    0.2760    0.6797    0.6551    0.1626    0.1190

EDU>> plot(x,y,'o',x,yp,'k:')
EDU>>

Elbarto

ixie99 · #3 $Old$ November 2nd, 2009, 04:41 PM

Quote:

Originally Posted by elbarto $View Post$

Try polyfit and polyeval.

Code:

EDU>> x = sort(rand(1,5))

x =

    0.4456    0.4898    0.6463    0.7094    0.7547

EDU>> y = rand(1,5)

y =

    0.2760    0.6797    0.6551    0.1626    0.1190

EDU>> p = polyfit(x,y,4)

p =

  1.0e+003 *

    1.0569   -2.3779    1.9403   -0.6777    0.0857

EDU>> yp=polyval(p,x)

yp =

    0.2760    0.6797    0.6551    0.1626    0.1190

EDU>> plot(x,y,'o',x,yp,'k:')
EDU>>

Elbarto

Thank you for your help. I understand most of the polyfit and polyval code. What troubles me the most right now is
"Using an appropriate range of signal to noise ratios, i need to construct the signal for 10 different signal to noise ratios ranging from 10:1 to 0.1:1. I need to generate polynomial fits to the data ranging from order 1 to order 4."
I am not sure how to construct the signal and i am lost with the signal to noise ratio.

elbarto · #4 $Old$ November 2nd, 2009, 05:42 PM

I cant help you out much there I am afraid. Do you have the signal processing toolbox installed?

Code:

EDU>> ver

will tell you what toolboxes you have available.

Something on this page might be of use Signal Processing Toolbox

Elbarto

CaptainBlack · #5 $Old$ November 2nd, 2009, 09:40 PM

Quote:

Originally Posted by ixie99 $View Post$

Hi all. I am new here. I though this would be a good place to ask for some help with my Matlab assignment, because i am completely lost.

I need to explore the detection of a signal in the presence of noise, the signal being the first 90 degrees of a sine wave. I need to analyze how good a fit of a polynomial with order and degree as a function increasing noise.

Using an appropriate range of signal to noise ratios, i need to construct the signal for 10 different signal to noise ratios ranging from 10:1 to 0.1:1. I need to generate polynomial fits to the data ranging from order 1 to order 4.

The last part is to calculate the residuals for each combination and demonstrate when a higher order fit is better than a linear fit.

I really need help on this. Any amount will do.
Thanks

You need to tell us something more about the noise, like its bandwidth .. Or about the signal sampling, or ...

CB

gaby99 · #6 $Old$ November 10th, 2009, 09:28 AM

Is that an online program. I nver heard of it

elbarto · #7 $Old$ November 10th, 2009, 04:28 PM

Is what an online program?

If you are talking about the signal processing toolbox then it is an optional feature of MATLAB that comes with the recent student versions. I am using 2008b an I have this toolbox available on my install.

Elbarto