Dear list,
is it possible to model user-defined correlation structure based on the
distance rather on the neighborhood?
I mean how to model
err[1:N]~dmnorm(err.mu[1:N],err.tau[1:N,1:N])
with hight-dimensional data (N about100 or even 1000) and eps.tau
depending on some random parameter, for example correlation or
variogram, to be fitted from the data.
The main issue is how to avoid the inversion of huge matrix and make use
of the fact that matrix is is sparce; "traditional" way of
err.tau[1:N,1:N]=inverse(err.sigma[1:N,1:N]);
err.sigma[i,j]=fun(distance(object_i,object_j),param) and
param~dnorm(par.mean,par.tau) with const r.mean,r.tau
takes too long time. What is defined in the geoBUGS (car.normal) is
based rather on the lattice model, but I need "geostatistical" one,
defined for the random field ...
Best regards,
Anatoly Saveliev
Kazan State University, Russia
-------------------------------------------------------------------
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
|