Exponential distributed random values

Hello,

I read in my script the following for creating exponential distributed random values:

n=50; %number of summands

n_exp=1000; %number of experiments

mu=rand(n,1)+0.5; %random means between 0.5 and 1.5

sigma=mu; %for exponential distribution

x=exprnd(repmat(mu,1,n_exp),n,n_exp); &exponential distributed random values for every experiment

n and n_exp are defining the size of the array but can anyone explain why I have to use repmat(mu,1,n_exp)?

Answer 1

[John D'Errico]

You need to do that since your exponential rate parameter varies. If it were constant, this would not be necessary. The first argument would just then be a scalar.