Search All of the Math Forum:
Views expressed in these public forums are not endorsed by
NCTM or The Math Forum.



How to generate bivariate time series data with given dependence structure?
Posted:
Nov 30, 2006 4:20 PM


I try to generate the bivariate time series data. Let's say Yt = A + B*Yt1 + Vt (1) Yt=(Y1t, Y2t) is the bivariate time series. If the dependence between Y1t and Y2t are measured by a copula function. H(Y1t, Y2t)=C(F(Y1t),G(Y2t)), where H is the joint distribution function, F anf G are the marginal distribution function. C are the copula function.
My question is: How to get a series Y1t has the given joint distribution H(Y1t, Y2t)=C(F(Y1t),G(Y2t)) and Yt follow the process of equation (1)? Thanks for help.
Susan



