spatial.gev.bma | R Documentation |
This is the main function of the spatial.gev.bma package. It runs an MCMC to sample the posterior distribution of a spatial GEV model that takes a number of covariates for linear interactions as well as a spatial component to model dependent overdispersion.
spatial.gev.bma(Y.list, X.all, S, n.reps, prior.user = NULL,
full = FALSE, fixed.xi = NULL, print.every = 0)
Y.list |
This is a list of length n, where n is the total number of sites you have data for. Each element of the list is a vector of observations at that site. |
X.all |
This is an n by p matrix where n is the total number of sites and p is the total number of variables under consideration. Please supply your own constant in the first column. Note that this model implicitly assumes that at a given site, the covariates are constant throughout the observation period. Email me if you need something more flexible. |
S |
This is an n by 2 matrix with the coordinates of each of your n sites. |
n.reps |
The total number of repititions (i.e. includes burn in) that you would like to run the MCMC. |
prior.user |
A list containing any priors that the user would like to specify. |
full |
A boolean indicating if you'd like to force all variables to be included (TRUE) or whether you want to perform BMA (FALSE) |
fixed.xi |
If set to NULL, the shape parameter is estimated like the location and scale. If a numeric value is provided, the shape parameter is fixed at this level. |
print.every |
A simple tool to tell you how many iterations you've run. If you set it to 0 then no progress reports will be given. If it is a positive number, say x, then every x iterations it will tell you the total number of iterations. |
TAU |
An n by reps by 3 array of the states of the random effects at each iteration |
THETA |
An reps by p by 3 array of the states of the linear terms at each iteration |
ALPHA |
A reps by 3 matrix of states of the alpha term at each iteration |
LAMBDA |
A reps by 3 matrix of states of the lambda term at each iteration |
Alex Lenkoski <alex@nr.no>
data(norway)
attach(norway)
##To replicate our results, change 2 to 2e5 below
a <- spatial.gev.bma(Y.list,X,S,2)
Add the following code to your website.
For more information on customizing the embed code, read Embedding Snippets.