Dear WinBUGS users,
Data consists of a time series of 25 days where the coefficents (lambda, phi)
are estimated for each hour. I'm trying to run the next normal hierarchical
model
model{
for (t in 2:T){
output[t,1:2]~dmnorm(mu[t,1:2],V[1:2,1:2])
mu[t,1]<-lambda[hour[t]]*output[t-1,2];
mu[t,2]<-phi[hour[t]]*output[t-1,2];
}
for(j in 1:198){
lambda[j]~dnorm(lambdamean,lambdaprecision)
phi[j]~dnorm(phimean,phiprecision)
}
lambdamean~dnorm(0.5,0.0000001);
lambdaprecision~dgamma(1, 0.000001)
phimean~dnorm(0.5,0.0000001);
phiprecision~dgamma(1, 0.000001)
V[1:2,1:2]~dwish(A[,],f1)
}
The coefficients lambda[hour] are estimated without problems. However, any
problem of convergence occurs for the coefficients phi[hour]. The estimation of
all the coefficients is the same, and the autocorrelations for any lag are near
to 1. If a try to estimate the coeffients each 30 minutes, the problems of
convergence occurs in both parameters. I proved varing the thin but the results
were the same. Can anybody help me to improve the convergence of the model?
Thanks in advance,
Miguel
--
Miguel A. Negrín Hernández
Modulo D. D-4.08
Department of Quantitative Methods
Facc. Ciencias Económicas y Empresariales
University of Las Palmas de Gran Canaria
35017. Las Palmas de G.C. Spain
Tfo.: (+34) 928 45 82 19
Fax: (+34) 928 45 82 25
E-mail: [log in to unmask]
-------------------------------------------------------------------
This list is for discussion of modelling issues and the BUGS software.
For help with crashes and error messages, first mail [log in to unmask]
To mail the BUGS list, mail to [log in to unmask]
Before mailing, please check the archive at www.jiscmail.ac.uk/lists/bugs.html
Please do not mail attachments to the list.
To leave the BUGS list, send LEAVE BUGS to [log in to unmask]
If this fails, mail [log in to unmask], NOT the whole list
|