plot2Q: Function to produce image plot of 2-dimensional data modeled... In CARrampsOcl: Reparameterized and marginalized posterior sampling for conditional autoregressive models, OpenCL implementation

Description

Function to produce image plot of 2-dimensional data modeled with 2 separate structure matrices.

Usage

 1 2 plot2Q(objname, numcols = 64, col = rev(terrain.colors(numcols)), rev.inds = c(FALSE, FALSE))

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

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 posterior distributions of the corresponding random effects.

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 posterior distributions of the corresponding random effects.

Kate Cowles

Examples

 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 # load data data(iowaSW06) # construct structure matrix Q1<- makeRW2Q(33) # for rows Q2<- makeRW2Q(24) # for columns # dimensions of Q1, Q2, in that order na<- nrow(Q1) nb<- nrow(Q2) Q <- list( list(type="Gen",content=Q1), list(type="Gen",content=Q2) ) # construct the design matrix with with as many columns as there are # in null space of kronecker prod of Q's X2 <- cbind( rep(1,nb), 1:nb) X1 <- cbind( rep(1,na), 1:na) X <- kronecker( X2, X1) # parameters of gamma prior densities on tausqy, tausqphi, tausqphi alpha2 = beta2 <- c(.1, .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 plot2Q( output, numcols=32, col = rev(terrain.colors(32)), rev.inds = c(FALSE, TRUE))

CARrampsOcl documentation built on May 2, 2019, 3:27 a.m.