transformation of VARMA model into its reduced form
[F,A,V,G]=theta2arma(theta,theta2mat)
* theta = (npx1) vector of parameters
* theta2mat = a string vector of instructions that transforms
* F = the MA part of the model
* A = the MA part of the model
* V = the variance of residuals
* G = the exogenous part of the model
(I + AR1.B + ... +ARp.B^p)(I + ARS1.B^s + ... + ARSps.B^ps.s) y(t) =
(G0 + G1.B + ... + Gt.B^l) u(t) + (I + MA1.B + ... + MAq.B^q)(I + MAS1.B^s + ... + MASqs.B^qs.s) a(t)
are transformed to:
F(B)y(t) = G(B)u(t) + A(B)e(t)