How can I add random noise using the rand function with the noise being uniformly distributed between 0 and 2?

If I have a new vector z=a1x^2 - a2y^2 + a3xy, how can I add random noise to all the elements in z using the function rand()? The noise needs to be uniformly distributed between 0 and 2.

Answer 1

[Amit]

z=a1*x.^2 - a2*y.^2 + a3*x.*y + 2*rand(size(x));

[Adrian]

Thank you so much. If I had already defined the vector before as z=a1*x.^2 - a2*y.^2 + a3*x.*y Can I just write: z = z + 2*rand(size(x)); ?

[Amit]

That would work.

[Adrian]

Thank you. I just had one more question. If I need to use the nlinfit() function to find the coefficients 1,2,3 so that so that the surface of z is the best fit to the noisy data, do you know how I could do that?

[Amit]

Didn't you already answered that? nlinfit()

For documentation regarding nlinfit: http://www.mathworks.com/help/stats/nlinfit.html

[Adrian]

Ok, thank you.