CARrampsOcl (version 0.1.4)

plotCAR1: Function to plot 2-dimensional data modeled using a single structure matrix.

Description

Function to plot 2-dimensional data modeled using a single structure matrix.

Usage

plotCAR1(objname, numcols = 64, col = rev(terrain.colors(numcols)), cardims, 
   rev.inds = c(FALSE, FALSE), 
   title = c("Raw data", "Estimated underlying truth"),
   sub = NULL)

Arguments

objname

name of output object produced by CARrampsOcl.fit

numcols

number of shades from the color palette to be used

col

color palette to be used in plotting; the default plots high values in green and low values in pink.

cardims

Two-vector specifying the number of rows and columns in the CAR1 structure matrix; same as dimension of CAR1 Q matrix provided to CARrampsOcl.fit.

rev.inds

Should the plotting indices on each two-dimensional plot be reversed? Setting rev.inds = c(TRUE,FALSE) flips the plots from left to right; rev.inds = c(FALSE,TRUE) turns the plots upside down.

title

Vector of character values; the titles of the two plots.

sub

subtitle for plots; vector of length equal to the number of pages of plots to be displayed

Value

• This function plots two two-dimensional plots side-by-side. The left plot is of the raw data input into the CARrampsOcl.fit function, and the right plot is of the estimated means of the of the posterior distributions of the corresponding random effects.

Details

This function plots two two-dimensional plots side-by-side. The left plot is of the raw data input into the CARrampsOcl.fit function, and the right plot is of the estimated means of the of the posterior distributions of the corresponding random effects.

Examples

# load data
data(iowaSW06)

Q <- list( list(type="CAR1",content=c(33,24)) )

# construct the design matrix with with as many columns as there are
# in null space of kronecker prod of Q's

X <-  matrix( rep(1,33*24), ncol=1)

# parameters of gamma prior densities on tausqy, tausqphi[1], tausqphi[2]
alpha2 = beta2 <- c(.1, .1)
# number of samples
nsamp = 100

#random seed
myseed = 314

output <- CARrampsOcl.fit(alpha=alpha2,
            beta=beta2, Q=Q, y=iowaSW06,  nsamp=nsamp,
            seed=myseed,
            fixed = FALSE, randeffs=TRUE, coefs=TRUE,designMat=X,
            mult= 50)

# plot the raw data and the posterior means of the site-specific random effects

plotCAR1( output, numcols=32, col = rev(terrain.colors(32)),
    cardims = c(33 ,24 ), rev.inds = c(FALSE, TRUE))