It would be really nice if someone could present an example for the bivariate normal with ghk function!
(Perhaps with some info on the actual purpose of this function)

Thanks
Pindar

Am 25.10.2012 11:32, schrieb Pindar:
Hi there,

recently I wrote a UDF in order to calculate and plot the bivariate normal distribution.
Now I found the ghk function and tried to replicate the results, but unfortunately in vain.
I think that I miss-specify the second and third argument. How is a correct?

<hansl>
matrix U=mnormal(400,2)
matrix A =seq(0,400)/-100 #lower bound   
A=A'~A'
matrix B =seq(0,400)/100 #upper bound   
B=B'~B'
matrix vcv=unvech({1;0;2})
matrix f = ghk(cholesky(vcv),A,B,U')

nulldata 400 --preserve
series x1=U[,1]
series x2=U[,2]
series fxy=f[1:400]
<hansl>

Thanks in advance
Leon Unger