/***********  cauchy.gau  *******************************
Author: H. D. Vinod, Econ Dept, Fordham Univ., Bronx, NY
        Director: Fordham Institute for Ethics and Economic Policy
        E-mail: vinod AT fordham.edu
        Web page  www.fordham.edu/economics/vinod
Date: July 12, 1996

Create cauchy distributed random variables
   standardized to have zero mean and unit variance, numerically
   even though it really does not have mean or variance.
INPUT: n=no of observations or no. of rows
       p=no. of columns
       seed= seed for generating uniform r.v.
OUTPUT: x which is n X p
********************************************************/
proc (1)=cauchy(n,p,seed);
        local e,x,y,i,seed1;
x=zeros(n,p);
i=1;do while i<=p;
        seed1=seed+2*i;
        e = rndus(n,1,seed1);
        y = tan(pi*(e-0.5));
        x[.,i]=(y-meanc(y))/stdc(y);
        i=i+1;
endo;
retp(x);
endp;