VerifytheWishartevolutiondistributiontheoryofSection10.4.8inthe followinggeneralsetting. Theq qprecisionmatrix hastheWishartdistribution W(hA) for somedegreesof freedomh=n+q 1wheren>0sothath> q 1 andwhereAisthe inversesum-of-squaresmatrix.TheBartlett decomposition,oftenusedforsimulationofWishartmatricesaswellas theoreticaldevelopments (e.g.,Odell
Question:
VerifytheWishartevolutiondistributiontheoryofSection10.4.8inthe followinggeneralsetting.
Theq qprecisionmatrix hastheWishartdistribution W(hA)
for somedegreesof freedomh=n+q 1wheren>0sothath>
q 1 andwhereAisthe inversesum-of-squaresmatrix.TheBartlett decomposition,oftenusedforsimulationofWishartmatricesaswellas theoreticaldevelopments (e.g.,Odell andFeiveson1966), is that =
PUUPwherePistheuppertriangularCholeskycomponentofAso thatA=PP,and U=
u11 u12 u13 u1q 0 u22 u23 u2q 0 0 u33 u3q 0 0 0 uqq wherethenonzeroentriesareindependentrandomquantitieswithuij N(01) for 1 i
1) 2) forsomeconstants i (01) for i=1:q The i aremutually independent,andindependentofthenormaluij Showthat, for i=1 : q i 2 i(h i+1) independentlyover iand independentlyoftheupper,o-diagonalelementsofU Deducethat W(k b 1A)when i=(k+i 1) (h+i 1) for eachi=1 q andwhere0 Identifythevaluesof(k b)andthe i ineachofthesespecialcases.
Step by Step Answer:
Time Series Modeling Computation And Inference
ISBN: 9781498747028
2nd Edition
Authors: Raquel Prado, Marco A. R. Ferreira, Mike West