Hi. How can i add 20% Gaussian noise to some fixed value vector.

Suppose i have a fixed value vector [1.05;0.2561;.3321]). I want to add 20% Gaussain noise to this vector such that this vector become the mean of the Gaussian distribution.

I hope that the question is clear

Thanks

[John D'Errico]

No, it is not clear. 20% noise suggests that you do not actually want classical additive Gaussian noise, but proportional noise.

[Fiaz Ahmad]

Hi Thanks John.

Actually i have the vector give above. Now what i want is to take this vector as mean and vary this randomly. The random variation is to be Gaussian 20% around the mean.

Answer 1

[Rick Rosson]

   m = [ 1.05 ; 0.2561 ; 0.3321 ];
   s = 0.2 * m ;
   N = 1000;

   x =  ones(N,1) * m'  +  randn(N,3) * diag(s);

[Fiaz Ahmad]

Thanks Rick it solved the problem