plot.pgls.R2AIC {mmodely} | R Documentation |
Plot (R2 vs AIC) results of a collection of fit PGLS models
Description
Plots a single panel of R^2 versus AIC, using versions of your choosing.
Usage
## S3 method for class 'pgls.R2AIC'
plot(x,
bests=bestBy(x, by=c('n','q','qXn','rwGsm')[4], best=c('AICc','R2.adj')[1],
inverse=c(FALSE,TRUE)[1]),bcl=rgb(1,1,1,maxColorValue=3,alpha=1), nx=2,
model.as.title='', ...)
Arguments
x |
a PGLSi[teration]$optim [optimization] table |
bests |
a list of the best PGLS models grouped by variable count and sorted by some metric (e.g. adjusted R2) |
bcl |
background color of plot point |
nx |
point size expansion factor to multiply against sample size ratio (this model to max of all models) |
model.as.title |
uses model.1ln.report to create a short character string of the "best" model results as a title |
... |
other parameters passed to 'plot' |
Value
a plot of R2 versus AIC of many PGLS models
Examples
data.path <- system.file("extdata","primate-example.data.csv", package="mmodely")
data <- read.csv(data.path, row.names=1)
pvs <- names(data[3:6])
data$gn_sp <- rownames(data)
tree.path <- system.file("extdata","primate-springer.2012.tre", package="mmodely")
phyl <- ape::read.tree(tree.path)[[5]]
mods <- get.model.combos(predictor.vars=pvs, outcome.var='OC', min.q=2)
# sprinkle in some missing data so as to make model selection more interesting
for(pv in pvs){ data[sample(x=1:nrow(data),size=2),pv] <- NA}
PGLSi <- pgls.iter(models=mods, phylo=phyl, df=data, k=1,l=1,d=1)
plot.pgls.R2AIC(PGLSi$optim) # find the lowest AIC within each q by n sized sub-dataset
[Package mmodely version 0.2.5 Index]