Stephen Montgomery-Smith wrote: > comtech wrote: > > Hi all, > > > > Suppose I have a vector random process call it Xt, > > > > where Xt is a vector of [X1t, X2t, X3t, ... Xm_t], so there are m > > sub-processes inside this vector random process. > > > > What are the conditions that the whole vector process is a Gaussian > > vector process? > > > > Maybe I just need to decide whether each individual process is Gaussian > > or not? > > > > Does individual Gaussianity of each individidual process <=> the whole > > vector process is Gaussian? > > > > No. For example, > > X1 = g > X2 = s g > > where g is a centered Gaussian and s takes the value plus or minus 1 > with prob 1/2, and g and s are independednt r.v.
Okay, so what's the precise statement for judging a vector process is a Gaussian vector process?