spTDyn (version 2.0.3)

decay: Choice for sampling spatial decay parameter \(\phi\).

Description

This function initialises the sampling method for the spatial decay parameter \(\phi\).

Usage

decay(distribution=Gamm(a=2,b=1), tuning=NULL, npoints=NULL, value=NULL)

Arguments

distribution

Prior distribution for \(\phi\). Currently available methods are, Gamm(a,b) and Unif(low,up). One can also used "FIXED" value for \(\phi\) parameter.

tuning

If the Gamma prior distribution is used then we need to define the tuning parameter for sampling \(\phi\). The tuning is the standard deviation for the normal proposal distribution of the random-walk Metropolis algorithm used to sample \(\phi\) on the log-scale.

npoints

If Unif distribution is used then need to define the number of segments for the range of limits by npoints. Default value is 5.

value

If distribution="FIXED" type is used then need to define the value for \(\phi\). The default value is 3/dmax where dmax is the maximum distance between the fitting sites provided by coords.

See Also

GibbsDyn.

Examples

## 

# input for random-walk Metropolis within Gibbs 
# sampling for phi parameter
spatial.decay<-decay(distribution=Gamm(2,1), tuning=0.08)

# input for discrete sampling of phi parameter 
# with uniform prior distribution
spatial.decay<-decay(distribution=Unif(0.01,0.02),npoints=5)

# input for spatial decay if FIXED is used
spatial.decay<-decay(distribution="FIXED", value=0.01)

##