#' Process mixing model output from JAGS
#'
#' \code{output_JAGS} processes the mixing model output, prints and saves (in the
#' working directory):
#' \itemize{
#' \item diagnostics
#' \item summary statistics
#' \item posterior density plots
#' \item pairs plot
#' \item trace/XY plots
#' }
#'
#' @param jags.1 rjags model object, output from \code{\link{run_model}} function
#' @param mix output from \code{\link{load_mix_data}}
#' @param source output from \code{\link{load_source_data}}
#' @param output_options list containing options for plots and saving:
#' \itemize{
#' \item \code{summary_save}: Save the summary statistics as a txt file? Default = \code{TRUE}
#' \item \code{summary_name}: Summary statistics file name (.txt will be appended). Default = \code{"summary_statistics"}
#' \item \code{sup_post}: Suppress posterior density plot output in R? Default = \code{FALSE}
#' \item \code{plot_post_save_pdf}: Save posterior density plots as pdfs? Default = \code{TRUE}
#' \item \code{plot_post_name}: Posterior plot file name(s) (.pdf/.png will be appended) Default = \code{"posterior_density"}
#' \item \code{sup_pairs}: Suppress pairs plot output in R? Default = \code{FALSE}
#' \item \code{plot_pairs_save_pdf}: Save pairs plot as pdf? Default = \code{TRUE}
#' \item \code{plot_pairs_name}: Pairs plot file name (.pdf/.png will be appended) Default = \code{"pairs_plot"}
#' \item \code{sup_xy}: Suppress xy/trace plot output in R? Default = \code{TRUE}
#' \item \code{plot_xy_save_pdf}: Save xy/trace plot as pdf? Default = \code{FALSE}
#' \item \code{plot_xy_name}: XY/trace plot file name (.pdf/.png will be appended) Default = \code{"xy_plot"}
#' \item \code{gelman}: Calculate Gelman-Rubin diagnostic test? Default = \code{TRUE}
#' \item \code{heidel}: Calculate Heidelberg-Welch diagnostic test? Default = \code{FALSE}
#' \item \code{geweke}: Calculate Geweke diagnostic test? Default = \code{TRUE}
#' \item \code{diag_save}: Save the diagnostics as a .txt file? Default = \code{TRUE}
#' \item \code{diag_name}: Diagnostics file name (.txt will be appended) Default = \code{"diagnostics"}
#' \item \code{indiv_effect}: artifact, set to FALSE
#' \item \code{plot_post_save_png}: Save posterior density plots as pngs? Default = \code{FALSE}
#' \item \code{plot_pairs_save_png}: Save pairs plot as png? Default = \code{FALSE}
#' \item \code{plot_xy_save_png}: Save xy/trace plot as png? Default = \code{FALSE}
#' \item \code{diag_save_ggmcmc}: Save ggmcmc diagnostics as pdf? Default = \code{TRUE}
#' }
#'
#' @return \code{p.both} -- only if 2 fixed effects OR 1 fixed + 1 random, otherwise \code{NULL}).
#'
#' \code{p.both} holds the MCMC chains for the estimated proportions at the different factor levels. Dimensions = [n.draws, f1.levels, f2.levels, n.sources].
#'
#' Calculated by combining the ilr offsets from global intercept:
#' ilr.both[,f1,f2,src] = ilr.global[,src] + ilr.fac1[,f1,src] + ilr.fac2[,f2,src]
#' And then transforming from ilr- to proportion-space.
#' @export
#'
output_JAGS <- function(jags.1, mix, source, output_options=list(
summary_save = TRUE, # Save the summary statistics as a txt file?
summary_name = "summary_statistics", # If yes, specify the base file name (.txt will be appended later)
sup_post = FALSE, # Suppress posterior density plot output in R?
plot_post_save_pdf = TRUE, # Save posterior density plots as pdfs?
plot_post_name = "posterior_density", # If yes, specify the base file name(s) (.pdf/.png will be appended later)
sup_pairs = FALSE, # Suppress pairs plot output in R?
plot_pairs_save_pdf = TRUE, # Save pairs plot as pdf?
plot_pairs_name = "pairs_plot", # If yes, specify the base file name (.pdf/.png will be appended later)
sup_xy = TRUE, # Suppress xy/trace plot output in R?
plot_xy_save_pdf = FALSE, # Save xy/trace plot as pdf?
plot_xy_name = "xy_plot", # If yes, specify the base file name (.pdf/.png will be appended later)
gelman = TRUE, # Calculate Gelman-Rubin diagnostic test?
heidel = FALSE, # Calculate Heidelberg-Welch diagnostic test?
geweke = TRUE, # Calculate Geweke diagnostic test?
diag_save = TRUE, # Save the diagnostics as a txt file?
diag_name = "diagnostics", # If yes, specify the base file name (.txt will be appended later)
indiv_effect = FALSE, # Is Individual a random effect in the model? (already specified)
plot_post_save_png = FALSE, # Save posterior density plots as pngs?
plot_pairs_save_png = FALSE, # Save pairs plot as png?
plot_xy_save_png = FALSE,
diag_save_ggmcmc = TRUE)){ # Save ggmcmc diagnostics as pdf?
mcmc.chains <- jags.1$BUGSoutput$n.chains
N <- mix$N
n.re <- mix$n.re
n.effects <- mix$n.effects
if(n.re==1){
random_effects <- ifelse(mix$FAC[[1]]$re,mix$FAC[[1]]$name,mix$FAC[[2]]$name)
}
if(n.re==2){
random_effects <- mix$factors
}
n.sources <- source$n.sources
source_names <- source$source_names
# p.global <- ilr.global <- ilr.fac1 <- ilr.fac2 <- fac1.sig <- fac2.sig <- NULL
# ind.sig <- ..scaled.. <- p.fac1 <- p.fac2 <- p.ind <- sources <- NULL
# R2jags::attach.jags(jags.1)
# jags1.mcmc <- coda::as.mcmc(jags.1)
as.mcmc.rjags <- function(x){
n.chains <- x$n.chains
sims <- x$sims.array
n.thin <- x$n.thin
if(n.chains==1) return(coda::mcmc(sims[, 1, ], thin=n.thin))
out <- vector("list", length=n.chains)
for (i in seq(n.chains)) out[[i]] <- coda::mcmc(sims[, i, ], thin=n.thin)
out <- coda::mcmc.list(out)
coda::varnames(out) <- dimnames(sims)[[3]]
return(out)
}
jags1.mcmc <- as.mcmc.rjags(jags.1$BUGSoutput)
n.draws <- length(jags.1$BUGSoutput$sims.list$p.global[,1])
# Post-processing for 2 FE or 1FE + 1RE
# calculate p.both = ilr.global + ilr.fac1 + ilr.fac2
if(mix$fere){
fac2_lookup <- list()
for(f1 in 1:mix$FAC[[1]]$levels){
fac2_lookup[[f1]] <- unique(mix$FAC[[2]]$values[which(mix$FAC[[1]]$values==f1)])
}
ilr.both <- array(NA,dim=c(n.draws,mix$FAC[[1]]$levels, mix$FAC[[2]]$levels, n.sources-1))
p.both <- array(NA,dim=c(n.draws,mix$FAC[[1]]$levels, mix$FAC[[2]]$levels, n.sources))
cross.both <- array(data=NA,dim=c(n.draws,mix$FAC[[1]]$levels, mix$FAC[[2]]$levels,n.sources,n.sources-1))
e <- matrix(rep(0,n.sources*(n.sources-1)),nrow=n.sources,ncol=(n.sources-1))
for(i in 1:(n.sources-1)){
e[,i] <- exp(c(rep(sqrt(1/(i*(i+1))),i),-sqrt(i/(i+1)),rep(0,n.sources-i-1)))
e[,i] <- e[,i]/sum(e[,i])
}
for(i in 1:n.draws){
for(f1 in 1:mix$FAC[[1]]$levels) {
for(f2 in fac2_lookup[[f1]]){
for(src in 1:(n.sources-1)) {
ilr.both[i,f1,f2,src] <- jags.1$BUGSoutput$sims.list$ilr.global[i,src] + jags.1$BUGSoutput$sims.list$ilr.fac1[i,f1,src] + jags.1$BUGSoutput$sims.list$ilr.fac2[i,f2,src];
cross.both[i,f1,f2,,src] <- (e[,src]^ilr.both[i,f1,f2,src])/sum(e[,src]^ilr.both[i,f1,f2,src]);
# ilr.both[,f1,f2,src] <- ilr.global[,src] + ilr.fac1[,f1,src] + ilr.fac2[,f2,src];
}
for(src in 1:n.sources) {
p.both[i,f1,f2,src] <- prod(cross.both[i,f1,f2,src,]);
}
p.both[i,f1,f2,] <- p.both[i,f1,f2,]/sum(p.both[i,f1,f2,]);
} # f2
} # f1
}
} # end fere
###########################################################################################
# XY/Trace Plots
###########################################################################################
# XY plots for p.global and factor SD's
if(!output_options[[9]]){ # if 'suppress XY plot' is NOT checked
# XY plot for p.global
dev.new()
print(lattice::xyplot(coda::as.mcmc(jags.1$BUGSoutput$sims.list$p.global),strip=lattice::strip.custom(factor.levels=source_names)))
# Save the xy p.global plot to file
if(output_options[[10]]){ # svalue(plot_xy_save_pdf)
mypath <- file.path(paste(getwd(),"/",output_options[[11]],"_diet_p.pdf",sep="")) # svalue(plot_xy_name)
dev.copy2pdf(file=mypath)
}
if(output_options[[20]]){ # svalue(plot_xy_save_png)
mypath <- file.path(paste(getwd(),"/",output_options[[11]],"_diet_p.png",sep="")) # svalue(plot_xy_name)
dev.copy(png,mypath)
}
# XY plot for the factor SDs
if(output_options[[17]]){ # include_indiv ('ind.sig' is in the model)
dev.new()
traceplot_labels <- rep("",length(random_effects)+1) # +1 because we need to add "Individual SD"
if(n.re > 0){
for(i in 1:length(random_effects)){
traceplot_labels[i] <- paste(random_effects[i]," SD",sep="")
}
}
traceplot_labels[length(random_effects)+1] <- "Individual SD"
if(n.re==2) print(lattice::xyplot(coda::as.mcmc(cbind(jags.1$BUGSoutput$sims.list$fac1.sig,jags.1$BUGSoutput$sims.list$fac2.sig,jags.1$BUGSoutput$sims.list$ind.sig)),strip=lattice::strip.custom(factor.levels=traceplot_labels)))
if(n.re==1){
if(mix$FAC[[1]]$re){
print(lattice::xyplot(coda::as.mcmc(cbind(jags.1$BUGSoutput$sims.list$fac1.sig,jags.1$BUGSoutput$sims.list$ind.sig)),strip=lattice::strip.custom(factor.levels=traceplot_labels)))
} else { # FAC 2 is the 1 random effect
print(lattice::xyplot(coda::as.mcmc(cbind(jags.1$BUGSoutput$sims.list$fac2.sig,jags.1$BUGSoutput$sims.list$ind.sig)),strip=lattice::strip.custom(factor.levels=traceplot_labels)))
}
}
if(n.re==0) print(lattice::xyplot(coda::as.mcmc(jags.1$BUGSoutput$sims.list$ind.sig),strip=lattice::strip.custom(factor.levels=traceplot_labels)))
} else { # Individual SD is not in the model (no 'ind.sig')
if(n.re > 0){
dev.new()
traceplot_labels <- rep("",length(random_effects))
for(i in 1:length(random_effects)) { traceplot_labels[i] <- paste(random_effects[i]," SD",sep="") }
if(n.re==2) print(lattice::xyplot(coda::as.mcmc(cbind(jags.1$BUGSoutput$sims.list$fac1.sig,jags.1$BUGSoutput$sims.list$fac2.sig)),strip=lattice::strip.custom(factor.levels=traceplot_labels)))
if(n.re==1){
if(mix$FAC[[1]]$re){
print(lattice::xyplot(coda::as.mcmc(cbind(jags.1$BUGSoutput$sims.list$fac1.sig)),strip=lattice::strip.custom(factor.levels=traceplot_labels)))
} else { # FAC 2 is the 1 random effect
print(lattice::xyplot(coda::as.mcmc(cbind(jags.1$BUGSoutput$sims.list$fac2.sig)),strip=lattice::strip.custom(factor.levels=traceplot_labels)))
}
}
}
}
# Save the xy factor SD plot to file
if(output_options[[10]]){ # svalue(plot_xy_save_pdf)
mypath <- file.path(paste(getwd(),"/",output_options[[11]],"_SD.pdf",sep="")) # svalue(plot_xy_name)
dev.copy2pdf(file=mypath)
}
if(output_options[[20]]){ # svalue(plot_xy_save_png)
mypath <- file.path(paste(getwd(),"/",output_options[[11]],"_SD.png",sep="")) # svalue(plot_xy_name)
dev.copy(png,mypath)
}
}
# Fancy pairs plot of p.global
# Contour plots in the upper right, histograms on the diagonal, correlation coefficients in the lower left
if(!output_options[[6]]){ # if 'suppress pairs plot' is NOT checked
dev.new()
# Function: panel.hist (from ?pairs)
# Purpose: creates histograms on the diagonal of the pairs plot matrix
panel.hist <- function(x, ...){
usr <- par("usr"); on.exit(par(usr), add=TRUE)
par(usr = c(usr[1:2], 0, 1.5) )
h <- hist(x, plot = FALSE)
breaks <- h$breaks; nB <- length(breaks)
y <- h$counts; y <- y/max(y)
rect(breaks[-nB], 0, breaks[-1], y, col='blue', xlim=c(0,1),...)
}
# Function: panel.cor (from http://personality-project.org/r/r.graphics.html)
# Purpose: prints correlation coefficients in the lower panel,
# scales text sizes to the correlation coefficient magnitudes
panel.cor <- function(x, y, digits=2, prefix="", cex.cor){
usr <- par("usr"); on.exit(par(usr), add=TRUE)
par(usr = c(0, 1, 0, 1))
r = (cor(x, y,use="pairwise"))
txt <- format(c(r, 0.123456789), digits=digits)[1]
txt <- paste(prefix, txt, sep="")
if(missing(cex.cor)) cex <- 0.8/strwidth(txt)
text(0.5, 0.5, txt, cex = cex * abs(r))
}
# Function: panel.contour (inspired by http://stats.stackexchange.com/questions/31726/scatterplot-with-contour-heat-overlay)
# Purpose: replaces scatterplots with colored contour plots
panel.contour <- function(x,y){
n.lines <- 4 # number of contour lines
my.cols <- rev(RColorBrewer::brewer.pal(n.lines, "RdYlBu")) # gets some pretty colors
z <- MASS::kde2d(x,y) # calculates the 2D kernel density that the contour function needs
contour(z, drawlabels=FALSE, nlevels=n.lines, col=my.cols, add=TRUE)
}
pairs(jags.1$BUGSoutput$sims.list$p.global, labels=source_names, diag.panel=panel.hist, lower.panel=panel.cor, upper.panel=panel.contour)
# Save the plot to file
if(output_options[[7]]){ # svalue(plot_pairs_save_pdf)
mypath <- file.path(paste(getwd(),"/",output_options[[8]],".pdf",sep="")) # svalue(plot_pairs_name)
dev.copy2pdf(file=mypath)
}
if(output_options[[19]]){ # svalue(plot_pairs_save_png)
mypath <- file.path(paste(getwd(),"/",output_options[[8]],".png",sep="")) # svalue(plot_pairs_name)
dev.copy(png,mypath)
}
}
######################################################################
# Posterior density plots
######################################################################
if(!output_options[[3]]){ # if 'suppress posterior plots' is NOT checked
n.draws <- length(jags.1$BUGSoutput$sims.list$p.global[,1]) # number of posterior draws
if(mix$n.fe == 0){ # only if there are no fixed effects, otherwise p.global is meaningless
# Posterior density plot for p.global
dev.new()
df <- data.frame(sources=rep(NA,n.draws*n.sources), x=rep(NA,n.draws*n.sources)) # create empty data frame
for(i in 1:n.sources){
df$x[seq(1+n.draws*(i-1),i*n.draws)] <- as.matrix(jags.1$BUGSoutput$sims.list$p.global[,i]) # fill in the p.global[i] values
df$sources[seq(1+n.draws*(i-1),i*n.draws)] <- rep(source_names[i],n.draws) # fill in the source names
}
my.title <- "Overall Population"
print(ggplot2::ggplot(df, ggplot2::aes(x=x, fill=sources, colour=sources)) +
ggplot2::geom_density(alpha=.3, ggplot2::aes(y=..scaled..)) +
ggplot2::theme_bw() +
ggplot2::xlab("Proportion of Diet") +
ggplot2::ylab("Scaled Posterior Density") +
ggplot2::xlim(0,1) +
ggplot2::labs(title = my.title) +
ggplot2::theme(legend.position=c(1,1), legend.justification=c(1,1), legend.title=ggplot2::element_blank()))
# Save the plot to file
if(output_options[[4]]){ # svalue(plot_post_save_pdf)
mypath <- file.path(paste(getwd(),"/",output_options[[5]],"_diet_p_global.pdf",sep="")) # svalue(plot_post_name)
dev.copy2pdf(file=mypath)
}
if(output_options[[18]]){ # svalue(plot_post_save_png)
mypath <- file.path(paste(getwd(),"/",output_options[[5]],"_diet_p_global.png",sep="")) # svalue(plot_post_name)
dev.copy(png,mypath)
}
}
if(n.effects >= 1 & mix$n.fe != 2){
# Posterior density plots for p.fac1's
for(f1 in 1:mix$FAC[[1]]$levels){ # formerly factor1_levels
dev.new()
df <- data.frame(sources=rep(NA,n.draws*n.sources), x=rep(NA,n.draws*n.sources)) # create empty data frame
for(src in 1:n.sources){
df$x[seq(1+n.draws*(src-1),src*n.draws)] <- as.matrix(jags.1$BUGSoutput$sims.list$p.fac1[,f1,src]) # fill in the p.fac1[f1] values
df$sources[seq(1+n.draws*(src-1),src*n.draws)] <- rep(source_names[src],n.draws) # fill in the source names
}
my.title <- mix$FAC[[1]]$labels[f1] # formerly factor1_names
print(ggplot2::ggplot(df, ggplot2::aes(x=x, fill=sources, colour=sources)) +
# geom_density(alpha=.3) +
ggplot2::geom_density(alpha=.3, ggplot2::aes(y=..scaled..)) +
ggplot2::xlim(0,1) +
ggplot2::theme_bw() +
ggplot2::xlab("Proportion of Diet") +
ggplot2::ylab("Scaled Posterior Density") +
ggplot2::labs(title = my.title) +
ggplot2::theme(legend.position=c(1,1), legend.justification=c(1,1), legend.title=ggplot2::element_blank()))
# Save the plot to file
if(output_options[[4]]){ # svalue(plot_post_save_pdf)
mypath <- file.path(paste(getwd(),"/",output_options[[5]],"_diet_p_",mix$FAC[[1]]$labels[f1],".pdf",sep="")) # svalue(plot_post_name), factor1_names
dev.copy2pdf(file=mypath)
}
if(output_options[[18]]){ # svalue(plot_post_save_png)
mypath <- file.path(paste(getwd(),"/",output_options[[5]],"_diet_p_",mix$FAC[[1]]$labels[f1],".png",sep="")) # svalue(plot_post_name), factor1_names
dev.copy(png,mypath)
}
} # end p.fac1 posterior plots
if(n.re==2){
# Posterior density plots for p.fac2's
for(f2 in 1:mix$FAC[[2]]$levels){ # formerly factor2_levels
dev.new()
df <- data.frame(sources=rep(NA,n.draws*n.sources), x=rep(NA,n.draws*n.sources)) # create empty data frame
for(src in 1:n.sources){
df$x[seq(1+n.draws*(src-1),src*n.draws)] <- as.matrix(jags.1$BUGSoutput$sims.list$p.fac2[,f2,src]) # fill in the p.fac2 values
df$sources[seq(1+n.draws*(src-1),src*n.draws)] <- rep(source_names[src],n.draws) # fill in the source names
}
#my.title <- paste(factor1_names[f1],", ", random_effects[2]," ",f2,sep="") # plot title (ex. "Region 1, Pack 3")
my.title <- mix$FAC[[2]]$labels[f2] # formerly factor2_names
print(ggplot2::ggplot(df, ggplot2::aes(x=x, fill=sources, colour=sources)) +
# geom_density(alpha=.3) +
ggplot2::geom_density(alpha=.3, ggplot2::aes(y=..scaled..)) +
ggplot2::theme_bw() +
ggplot2::xlim(0,1) +
ggplot2::xlab("Proportion of Diet") +
ggplot2::ylab("Scaled Posterior Density") +
ggplot2::labs(title = my.title) +
ggplot2::theme(legend.position=c(1,1), legend.justification=c(1,1), legend.title=ggplot2::element_blank()))
# Save the plot as a pdf file
if(output_options[[4]]){ # svalue(plot_post_save_pdf)
mypath <- file.path(paste(getwd(),"/",output_options[[5]],"_diet_p_",mix$FAC[[2]]$labels[f2],".pdf",sep="")) # svalue(plot_post_name), factor2_names
dev.copy2pdf(file=mypath)
}
if(output_options[[18]]){ # svalue(plot_post_save_png)
mypath <- file.path(paste(getwd(),"/",output_options[[5]],"_diet_p_",mix$FAC[[2]]$labels[f2],".png",sep="")) # svalue(plot_post_name), factor2_names
dev.copy(png,mypath)
}
}# end p.fac2 posterior plots
} # end if(n.re==2)
} # end if(n.effects >=1 & n.fe != 2)
# Posterior density plots for p.both (when 2 FE or 1FE + 1RE)
if(mix$fere){
for(f1 in 1:mix$FAC[[1]]$levels) {
for(f2 in fac2_lookup[[f1]]){
dev.new()
df <- data.frame(sources=rep(NA,n.draws*n.sources), x=rep(NA,n.draws*n.sources)) # create empty data frame
for(src in 1:n.sources){
df$x[seq(1+n.draws*(src-1),src*n.draws)] <- as.matrix(p.both[,f1,f2,src]) # fill in the p.both values
df$sources[seq(1+n.draws*(src-1),src*n.draws)] <- rep(source_names[src],n.draws) # fill in the source names
}
my.title <- paste(mix$FAC[[1]]$labels[f1],mix$FAC[[2]]$labels[f2],sep=" ") # formerly factor2_names
print(ggplot2::ggplot(df, ggplot2::aes(x=x, fill=sources, colour=sources)) +
ggplot2::geom_density(alpha=.3, ggplot2::aes(y=..scaled..)) +
ggplot2::theme_bw() +
ggplot2::xlim(0,1) +
ggplot2::xlab("Proportion of Diet") +
ggplot2::ylab("Scaled Posterior Density") +
ggplot2::labs(title = my.title) +
ggplot2::theme(legend.position=c(1,1), legend.justification=c(1,1), legend.title=ggplot2::element_blank()))
# Save the plot as a pdf file
if(output_options[[4]]){ # svalue(plot_post_save_pdf)
mypath <- file.path(paste(getwd(),"/",output_options[[5]],"_diet_p_",mix$FAC[[1]]$labels[f1],"_",mix$FAC[[2]]$labels[f2],".pdf",sep="")) # svalue(plot_post_name), factor2_names
dev.copy2pdf(file=mypath)
}
if(output_options[[18]]){ # svalue(plot_post_save_png)
mypath <- file.path(paste(getwd(),"/",output_options[[5]],"_diet_p_",mix$FAC[[1]]$labels[f1],"_",mix$FAC[[2]]$labels[f2],".png",sep="")) # svalue(plot_post_name), factor2_names
dev.copy(png,mypath)
}
} # f2
} # f1
}
# Posterior density plot for fac1.sig, fac2.sig, and ind.sig
if(n.re > 0 || output_options[[17]]){ # only have an SD posterior plot if we have Individual, Factor1, or Factor2 random effects)
dev.new()
n.re_ind <- n.re + as.numeric(output_options[[17]]) # this*n.draws will be the length of the plot data frame
level <- c()
x <- c()
if(output_options[[17]]){ # if Individual is in the model, add ind.sig to the SD plot
level <- c(level,rep("Individual SD",n.draws))
x <- c(x,jags.1$BUGSoutput$sims.list$ind.sig)
}
if(n.re==1){ # if Factor.1 is in the model, add fac1.sig to the SD plot
if(mix$FAC[[1]]$re){
level <- c(level,rep(paste(mix$FAC[[1]]$name," SD",sep=""),n.draws))
x <- c(x,jags.1$BUGSoutput$sims.list$fac1.sig)
} else { # FAC 2 is the random effect
level <- c(level,rep(paste(mix$FAC[[2]]$name," SD",sep=""),n.draws))
x <- c(x,jags.1$BUGSoutput$sims.list$fac2.sig)
}
}
if(n.re==2){ # if Factor.2 is in the model, add fac1.sig and fac2.sig to the SD plot
level <- c(level,rep(paste(random_effects[1]," SD",sep=""),n.draws), rep(paste(random_effects[2]," SD",sep=""),n.draws))
x <- c(x,jags.1$BUGSoutput$sims.list$fac1.sig,jags.1$BUGSoutput$sims.list$fac2.sig)
}
df2 <- data.frame(level=level, x=x) # create the SD plot data frame
print(ggplot2::ggplot(df2, ggplot2::aes(x=x, fill=level, colour=level)) +
# geom_density(alpha=.3) +
ggplot2::geom_density(alpha=.3) +
ggplot2::theme_bw() +
ggplot2::xlab(expression(sigma)) +
ggplot2::ylab("Posterior Density") +
ggplot2::theme(legend.position=c(1,1), legend.justification=c(1,1), legend.title=ggplot2::element_blank())) # + xlim(0,2)
# Save the plot to file
if(output_options[[4]]){ # svalue(plot_post_save_pdf)
mypath <- file.path(paste(getwd(),"/",output_options[[5]],"_SD.pdf",sep="")) # svalue(plot_post_name)
dev.copy2pdf(file=mypath)
}
if(output_options[[18]]){ # svalue(plot_post_save_png)
mypath <- file.path(paste(getwd(),"/",output_options[[5]],"_SD.png",sep="")) # svalue(plot_post_name)
dev.copy(png,mypath)
}
}
}
# Calculate the summary statistics for the variables we're interested in (p.global's and factor SD's, maybe p.ind's)
# We print them out later, at the very bottom
sig_labels <- NULL; ind_labels <- NULL; fac1_labels <- NULL; fac2_labels <- NULL; sig_stats <- NULL;
getQuant <- function(x) quantile(x,probs=c(.025,.05,.25,.5,.75,.95,.975))
getMeanSD <- function(x) cbind(round(apply(x,2,mean),3),round(apply(x,2,sd),3))
stats <- NULL
sig_stats <- NULL
sig_labels <- NULL
eps_stats <- NULL
eps_labels <- NULL
# print(mix)
# print(mix$n.fe)
if(mix$n.fe == 0){
global_quants <- t(round(apply(jags.1$BUGSoutput$sims.list$p.global,2,getQuant),3))
global_means <- getMeanSD(jags.1$BUGSoutput$sims.list$p.global)
stats <- cbind(global_means, global_quants)
global_labels <- rep(NA,n.sources)
for(src in 1:n.sources){
global_labels[src] <- paste("p.global.",source_names[src],sep="")
}
rownames(stats) <- global_labels
}
if(n.effects > 0 & mix$n.fe != 2){
fac1_quants <- as.matrix(reshape::cast(reshape2::melt(round(apply(jags.1$BUGSoutput$sims.list$p.fac1,c(2,3),getQuant),3)),Var3+Var2~Var1)[,-c(1,2)])
fac1_quants <- t(apply(fac1_quants,1,sort)) # BUG FIX 10/28/14, quantiles were out of order from cast/melt (thanks to Jason Waite)
fac1_means <- cbind(reshape2::melt(round(apply(jags.1$BUGSoutput$sims.list$p.fac1,c(2,3),mean),3))$value, reshape2::melt(round(apply(jags.1$BUGSoutput$sims.list$p.fac1,c(2,3),sd),3))$value)
fac1_stats <- cbind(fac1_means,fac1_quants)
fac1_labels <- rep(NA,mix$FAC[[1]]$levels*n.sources)
for(src in 1:n.sources){
for(f1 in 1:mix$FAC[[1]]$levels){
fac1_labels[mix$FAC[[1]]$levels*(src-1)+f1] <- paste("p.",mix$FAC[[1]]$labels[f1],".",source_names[src],sep="")
}
}
rownames(fac1_stats) <- fac1_labels
stats <- rbind(stats,fac1_stats)
if(mix$FAC[[1]]$re){
sig_stats <- cbind(getMeanSD(jags.1$BUGSoutput$sims.list$fac1.sig),t(round(apply(jags.1$BUGSoutput$sims.list$fac1.sig,2,getQuant),3)))
sig_labels <- paste(mix$FAC[[1]]$name,".SD",sep="")
}
}
if(n.re==2){
fac2_quants <- as.matrix(reshape::cast(reshape2::melt(round(apply(jags.1$BUGSoutput$sims.list$p.fac2,c(2,3),getQuant),3)),Var3+Var2~Var1)[,-c(1,2)])
fac2_quants <- t(apply(fac2_quants,1,sort)) # BUG FIX 10/28/14, quantiles were out of order from cast/melt (thanks to Jason Waite)
fac2_means <- cbind(reshape2::melt(round(apply(jags.1$BUGSoutput$sims.list$p.fac2,c(2,3),mean),3))$value, reshape2::melt(round(apply(jags.1$BUGSoutput$sims.list$p.fac2,c(2,3),sd),3))$value)
fac2_stats <- cbind(fac2_means,fac2_quants)
fac2_labels <- rep(NA,mix$FAC[[2]]$levels*n.sources)
for(src in 1:n.sources){
for(f2 in 1:mix$FAC[[2]]$levels){
fac2_labels[mix$FAC[[2]]$levels*(src-1)+f2] <- paste("p.",mix$FAC[[2]]$labels[f2],".",source_names[src],sep="")
}
}
rownames(fac2_stats) <- fac2_labels
stats <- rbind(stats,fac2_stats)
if(mix$FAC[[2]]$re){
sig_stats <- rbind(sig_stats,cbind(getMeanSD(jags.1$BUGSoutput$sims.list$fac2.sig),t(round(apply(jags.1$BUGSoutput$sims.list$fac2.sig,2,getQuant),3))))
sig_labels <- c(sig_labels,paste(mix$FAC[[2]]$name,".SD",sep=""))
}
}
if(mix$fere){
fac2_quants <- matrix(NA,nrow=n.sources*length(unlist(fac2_lookup)),ncol=7)
fac2_means <- matrix(NA,nrow=n.sources*length(unlist(fac2_lookup)),ncol=2)
fac2_labels <- rep(NA,n.sources*length(unlist(fac2_lookup)))
i <- 1
for(f1 in 1:mix$FAC[[1]]$levels) {
for(f2 in fac2_lookup[[f1]]){
for(src in 1:n.sources){
fac2_quants[i,] <- getQuant(p.both[,f1,f2,src])
fac2_means[i,] <- c(mean(p.both[,f1,f2,src]),sd(p.both[,f1,f2,src]))
fac2_labels[i] <- paste("p",mix$FAC[[1]]$labels[f1],mix$FAC[[2]]$labels[f2],source_names[src],sep=".")
i <- i+1
}
}
}
# fac2_quants <- as.matrix(cast(melt(round(apply(p.both,c(2,3,4),getQuant,na.rm=TRUE),3)),X4+X3+X2~X1)[,-c(1,2)])
# fac2_quants <- t(apply(fac2_quants,1,sort)) # BUG FIX 10/28/14, quantiles were out of order from cast/melt (thanks to Jason Waite)
# fac2_means <- cbind(melt(round(apply(p.fac2,c(2,3),mean),3))$value, melt(round(apply(p.fac2,c(2,3),sd),3))$value)
fac2_stats <- round(cbind(fac2_means,fac2_quants),3)
rownames(fac2_stats) <- fac2_labels
stats <- rbind(stats,fac2_stats)
if(mix$FAC[[2]]$re){
sig_stats <- rbind(sig_stats,cbind(getMeanSD(jags.1$BUGSoutput$sims.list$fac2.sig),t(round(apply(jags.1$BUGSoutput$sims.list$fac2.sig,2,getQuant),3))))
sig_labels <- c(sig_labels,paste(mix$FAC[[2]]$name,".SD",sep=""))
}
}
if(output_options[[17]]){ # include_indiv (if Individual is in the model)
ind_quants <- as.matrix(reshape::cast(reshape2::melt(round(apply(p.ind,c(2,3),getQuant),3)),X3+X2~X1)[,-c(1,2)])
ind_quants <- t(apply(ind_quants,1,sort)) # BUG FIX 10/28/14, quantiles were out of order from cast/melt (thanks to Jason Waite)
ind_means <- cbind(reshape2::melt(round(apply(p.ind,c(2,3),mean),3))$value, reshape2::melt(round(apply(p.ind,c(2,3),sd),3))$value)
ind_stats <- cbind(ind_means,ind_quants)
ind_labels <- rep(NA,N*n.sources)
for(src in 1:n.sources){
for(j in 1:N){
ind_labels[N*(src-1)+j] <- paste("p.Ind ",j,".",source_names[src],sep="")
}
}
sig_stats <- rbind(sig_stats,cbind(getMeanSD(jags.1$BUGSoutput$sims.list$ind.sig),t(round(apply(jags.1$BUGSoutput$sims.list$ind.sig,2,getQuant),3))))
sig_labels <- c(sig_labels,"Individual.SD")
rownames(ind_stats) <- ind_labels
stats <- rbind(stats, ind_stats)
}
# Add SD stats to the top of the summary
rownames(sig_stats) <- sig_labels
stats <- rbind(sig_stats,stats)
# Add epsilon (multiplicative error term) to stat summary
# Also plot posterior density
epsTF <- "resid.prop" %in% names(jags.1$BUGSoutput$sims.list)
if(epsTF){
eps_stats <- cbind(getMeanSD(jags.1$BUGSoutput$sims.list$resid.prop),t(round(apply(jags.1$BUGSoutput$sims.list$resid.prop,2,getQuant),3)))
eps_labels <- paste0("Epsilon.", 1:mix$n.iso)
rownames(eps_stats) <- eps_labels
stats <- rbind(eps_stats,stats)
# posterior plot
level <- c()
x <- c()
for(j in 1:mix$n.iso){
level <- c(level,rep(eps_labels[j], n.draws))
x <- c(x, jags.1$BUGSoutput$sims.list$resid.prop[,j])
}
df2 <- data.frame(level=level, x=x)
dev.new()
print(ggplot2::ggplot(df2, ggplot2::aes(x=x, fill=level, colour=level)) +
ggplot2::geom_density(alpha=.3) +
ggplot2::theme_bw() +
ggplot2::xlab(expression(epsilon)) +
ggplot2::ylab("Posterior Density") +
ggplot2::theme(legend.position=c(.95,.95), legend.justification=c(1,1), legend.title=ggplot2::element_blank())) # + xlim(0,2)
}
colnames(stats) <- c("Mean","SD","2.5%","5%","25%","50%","75%","95%","97.5%")
# Pack 1 stats only
#stats[grep("Pack 1",rownames(stats)),]
# Region stats only
#stats[grep("Region",rownames(stats)),]
# Region stats, by Region
# byVec <- function(x){ind <- NULL; for(i in 1:length(x)){ ind <- c(ind,grep(x[i],rownames(stats)))}; return(ind)}
# stats[byVec(mix$RE[[1]]$labels),]
# All means
# stats[,"Mean"]
# Region means only
# stats[byVec(mix$RE[[1]]$labels),"Mean"]
################################################################################
# Calulate diagnostics
################################################################################
# Get number of variables in the model
n.var <- coda::nvar(jags1.mcmc)
# Gelman-Rubin diagnostic
if(output_options[[12]]){ # if Gelman is checked
if(mcmc.chains == 1){
gelman <- "*** Error: Gelman diagnostic requires more than one chain ***"
}
if(mcmc.chains > 1){ # Gelman diagnostic requires more than one chain
# Gelman diagnostic, for when the multivariate Gelman fails (matrix not positive definite)
# Remove the test results for dummy/empty variables
gelman <- matrix(NA, nrow=n.var, ncol=2)
for (v in 1:coda::nvar(jags1.mcmc)) {
gelman[v,] <- coda::gelman.diag(jags1.mcmc[,v])$psrf
}
#gelman <- gelman[ind,]
colnames(gelman) <- c("Point est.","Upper C.I.")
rownames(gelman) <- coda::varnames(jags1.mcmc)
#rownames(gelman) <- c(sig_labels,global_labels,fac1_labels,fac2_labels,ind_labels)
gelman.all <- gelman[which(!is.nan(gelman[,1])),] # Remove dummy variables (show up as NA)
gelman_short <- gelman[order(gelman[,1],decreasing=T),]
if(n.var>10) gelman_short <- gelman_short[1:10,]
gelman_fail <- c(length(which(gelman[,1]>1.01)), length(which(gelman[,1]>1.05)), length(which(gelman[,1]>1.1)))
}
}
# Heidelberger and Welch's diagnostic
# Remove the test results for dummy/empty variables
if(output_options[[13]]){ # if Heidel is checked
heidel <- coda::heidel.diag(jags1.mcmc)
w <- which(!is.na(heidel[[1]][,"pvalue"])) # find all the non-dummy variables
heidel.all <- data.frame(matrix(NA,nrow=length(w),ncol=3*mcmc.chains)) # create empty data frame
colstring <- rep(NA,mcmc.chains*3) # vector of column names
for(i in 1:mcmc.chains){
heidel.tmp <- as.data.frame(heidel[[i]][w,c("stest","pvalue","htest")]) # stest, pvalue, and htest are the relevant statistics - get them
heidel.all[,(3*i-2):(3*i)] <- heidel.tmp
colstring[(3*i-2):(3*i)] <- c(paste("stest.",i,sep=""), paste("pval.",i,sep=""), paste("hwtest.",i,sep="")) # create the appropriate column names
}
#heidel.all <- heidel.all[ind,]
#rownames(heidel.all) <- c(sig_labels,global_labels,fac1_labels,fac2_labels,ind_labels)
rownames(heidel.all) <- coda::varnames(jags1.mcmc)[w]
colnames(heidel.all) <- colstring
heidel.all <- round(heidel.all,3)
heidel.all <- replace(heidel.all,heidel.all==0,"fail") # A normal call to 'heidel.diag' prints "fail" and "pass", for some reason they turn to 0's and 1's
heidel.all <- replace(heidel.all,heidel.all==1,"pass") # when you access the statistics directly. Here we turn the 0's and 1's back into "fail" and "pass"
# When the stationarity test fails, hwtest returns <NA>...change these NAs to 'fail'
heidel.all <- replace(heidel.all,is.na(heidel.all),"fail")
# Count the number of failures (2 tests per chain - 'stationarity' and 'half-width')
stest_fail <- rep(NA,mcmc.chains); hwtest_fail <- rep(NA,mcmc.chains)
for(i in 1:mcmc.chains){
stest_fail[i] <- sum(heidel.all[,3*i-2]=="fail")
hwtest_fail[i] <- sum(heidel.all[,3*i]=="fail")
}
heidel_fail <- rbind(stest_fail,hwtest_fail)
rownames(heidel_fail) <- c("Stationarity","Half-width")
colnames(heidel_fail) <- paste("Chain",1:mcmc.chains)
}
# Geweke diagnostic
# Remove the test results for dummy/empty variables
if(output_options[[14]]){ # if Geweke is checked
geweke <- coda::geweke.diag(jags1.mcmc)
geweke.all <- data.frame(matrix(NA,nrow=n.var,ncol=mcmc.chains)) # create empty data frame
colstring <- rep(NA,mcmc.chains) # vector of column names
for(i in 1:mcmc.chains){
geweke.tmp <- as.data.frame(geweke[[i]]$z) # get the relevant geweke statistics
geweke.all[,i] <- geweke.tmp
colstring[i] <- c(paste("chain",i,sep="")) # create the column names "chain1", "chain2", etc.
}
#geweke.all <- geweke.all[ind,]
#rownames(geweke.all) <- c(sig_labels,global_labels,fac1_labels,fac2_labels,ind_labels)
rownames(geweke.all) <- coda::varnames(jags1.mcmc)
colnames(geweke.all) <- colstring
geweke.all <- round(geweke.all,3)
w <- which(!is.nan(geweke[[1]]$z)) # find all the non-dummy variables
geweke.all <- geweke.all[w,]
geweke_fail <- matrix(NA,nrow=1,ncol=mcmc.chains)
for(i in 1:mcmc.chains){
geweke_fail[1,i] <- sum(abs(geweke.all[,i])>1.96)
}
colnames(geweke_fail) <- paste("Chain",1:mcmc.chains)
rownames(geweke_fail) <- "Geweke"
}
################################################################################
# Print diagnostics
################################################################################
if(output_options[[12]]){ # svalue(gelman)
cat("
################################################################################
# Gelman-Rubin Diagnostic
################################################################################
Generally the Gelman diagnostic should be < 1.05
",paste("Out of ",n.var," variables: ",gelman_fail[1]," > 1.01",sep=""),"
",paste(gelman_fail[2]," > 1.05",sep=""),"
",paste(gelman_fail[3]," > 1.1",sep=""),"
The worst variables are:
",sep="")
print(gelman_short)
#print(gelman)
if(output_options[[15]]){ # svalue(diag_save)
mypath <- file.path(paste(getwd(),"/",output_options[[16]],".txt",sep="")) # svalue(diag_name)
out <- capture.output(gelman)
out2 <- capture.output(gelman_short)
cat("
################################################################################
# Gelman-Rubin Diagnostic
################################################################################
Generally the Gelman diagnostic should be < 1.05
",paste("Out of ",n.var," variables: ",gelman_fail[1]," > 1.01",sep=""),"
",paste(gelman_fail[2]," > 1.05",sep=""),"
",paste(gelman_fail[3]," > 1.1",sep=""),"
The worst variables are:
",out2,"
And here are the Gelman diagnostics for all variables:
",out,sep="\n", file=mypath, append=FALSE)
} # end save Gelman
} # end Gelman printout
if(output_options[[13]]){ # svalue(heidel)
cat("
################################################################################
# Heidelberger and Welch Diagnostic
################################################################################
A few failures is normal and acceptable...
Number of failures in each chain (out of ",n.var," variables):
",sep="")
print(heidel_fail)
#print(heidel.all)
if(output_options[[15]]){ # svalue(diag_save)
mypath <- file.path(paste(getwd(),"/",output_options[[16]],".txt",sep="")) # svalue(diag_name)
out <- capture.output(heidel.all)
out2 <- capture.output(heidel_fail)
cat("
################################################################################
# Heidelberger and Welch Diagnostic
################################################################################
A few failures is normal and acceptable...
Number of failures in each chain (out of ",n.var," variables):
",out2,"
And here are the Heidelberger-Welch diagnostics for all variables:
",out,sep="\n", file=mypath, append=output_options[[12]]) # svalue(gelman)
} # end save Heidel
} # end Heidel printout
if(output_options[[14]]){ # svalue(geweke)
cat("
################################################################################
# Geweke Diagnostic
################################################################################
The Geweke diagnostic is a standard z-score, so we'd expect 5% to be outside +/-1.96
Number of variables outside +/-1.96 in each chain (out of ",n.var,"):
",sep="")
print(geweke_fail)
#print(geweke.all)
if(output_options[[15]]){ # svalue(diag_save)
mypath <- file.path(paste(getwd(),"/",output_options[[16]],".txt",sep="")) # svalue(diag_name)
out <- capture.output(geweke.all)
out2 <- capture.output(geweke_fail)
cat("
################################################################################
# Geweke Diagnostic
################################################################################
The Geweke diagnostic is a standard z-score, so we'd expect 5% to be outside +/-1.96
Number of variables outside +/-1.96 in each chain (out of ",n.var,"):
",out2,"
And here are the Geweke diagnostics for all variables:
",out,sep="\n", file=mypath, append=output_options[[12]]||output_options[[13]]) # svalue(gelman) || svalue(heidel)
} # end Geweke save
} # end Geweke printout
DIC <- jags.1$BUGSoutput$DIC
cat("
################################################################################
# Summary Statistics
################################################################################
DIC = ",DIC,sep="")
out1 <- capture.output(stats)
cat("
",out1,sep="\n")
if(output_options[[1]]){ # svalue(summary_save)
mypath <- file.path(paste(getwd(),"/",output_options[[2]],".txt",sep="")) # svalue(summary_name)
cat("
#################################################################
# Summary Statistics
#################################################################
DIC = ",DIC,sep="", file=mypath, append=FALSE)
cat("
",out1,sep="\n", file=mypath, append=TRUE)
}
# Plot any continuous effects
if(mix$n.ce > 0){
plot_continuous_var(jags.1,mix,source,output_options)
}
# Use ggmcmc package to create diagnostic plots
if(!is.null(output_options$diag_save_ggmcmc)) if(output_options$diag_save_ggmcmc){
diag_filename <- paste(getwd(),"/",output_options$diag_name,".pdf",sep="")
ggmcmc::ggmcmc(ggmcmc::ggs(jags1.mcmc), file=diag_filename, plot=c("Rhat","geweke","density","traceplot","running","autocorrelation","crosscorrelation"))
}
# Return p.both if 2 FE or 1FE + 1RE
if(mix$fere){
return(p.both)
} else return(NULL) # otherwise return nothing
} # end function output_JAGS
Add the following code to your website.
For more information on customizing the embed code, read Embedding Snippets.