density.mcmcSTmodel: Kernel Density Estimation for an 'mcmcSTmodel' Object

In SpatioTemporal: Spatio-Temporal Model Estimation

Description

density method for class mcmcSTmodel.

Usage

## S3 method for class 'mcmcSTmodel'
density(x, BurnIn = 0, estSTmodel = NULL, ...)

Arguments

x	mcmcSTmodel object
BurnIn	Number of initial points to ignore.
estSTmodel	Either a estimateSTmodel object from estimate.STmodel or a matrix with parameter-estimates and standard deviations, such as the output from coef.estimateSTmodel. If given as a matrix, it should have columns named "par" and "sd", and rows named after the parameters.
...	Additional parameters passed to density.

Details

Computes kernel density estimates for the MCMC-parameters; as well as approximate Gaussian densities based on the Fischer-information.

Value

List containing density estimate and Gaussian densities for all model parameters.

Author(s)

Johan Lindstrom

See Also

Other mcmcSTmodel methods: MCMC.STmodel, plot.density.mcmcSTmodel, plot.mcmcSTmodel, print.mcmcSTmodel, print.summary.mcmcSTmodel, summary.mcmcSTmodel

Examples

##load estimation results
data(est.mesa.model)
##and MCMC results instead
data(MCMC.mesa.model)

##compute density estimates for the results, and use the Gaussian approximation
##based on Fischer information as reference.
dens <- density(MCMC.mesa.model, estSTmodel=est.mesa.model)

##all the estimated densities
str(dens,1)

##or results for one paramter
dens[[1]]

##plot density functions
plot(dens)
##for a different paramter, along with Gaussian approx
plot(dens, 3, norm.col="red")

##all covariance parameters
par(mfrow=c(3,3),mar=c(4,4,2.5,.5))
for(i in 9:17){
  plot(dens, i, norm.col="red")
}

Example output

Loading required package: Matrix
List of 19
 $ gamma.lax.conc.1500            :List of 2
 $ alpha.const.(Intercept)        :List of 2
 $ alpha.const.log10.m.to.a1      :List of 2
 $ alpha.const.s2000.pop.div.10000:List of 2
 $ alpha.const.km.to.coast        :List of 2
 $ alpha.V1.(Intercept)           :List of 2
 $ alpha.V1.km.to.coast           :List of 2
 $ alpha.V2.(Intercept)           :List of 2
 $ alpha.V2.km.to.coast           :List of 2
 $ log.range.const.exp            :List of 2
 $ log.sill.const.exp             :List of 2
 $ log.range.V1.exp               :List of 2
 $ log.sill.V1.exp                :List of 2
 $ log.range.V2.exp               :List of 2
 $ log.sill.V2.exp                :List of 2
 $ nu.log.range.exp               :List of 2
 $ nu.log.sill.exp                :List of 2
 $ nu.log.nugget.(Intercept).exp  :List of 2
 $ nu.log.nugget.typeFIXED.exp    :List of 2
 - attr(*, "class")= chr "density.mcmcSTmodel"
$density

Call:
	density.default(x = newX[, i])

Data: newX[, i] (2500 obs.);	Bandwidth 'bw' = 0.0004833

       x                   y            
 Min.   :-0.008585   Min.   :  0.00747  
 1st Qu.:-0.003368   1st Qu.:  6.48915  
 Median : 0.001849   Median : 24.59657  
 Mean   : 0.001849   Mean   : 47.87467  
 3rd Qu.: 0.007066   3rd Qu.: 85.99468  
 Max.   : 0.012282   Max.   :156.83253  

$approx.gauss
$approx.gauss$mean
[1] 0.0008978266

$approx.gauss$sd
[1] 0.002694145

SpatioTemporal documentation built on May 2, 2019, 8:49 a.m.