ggplot.evmOpt: Diagnostic plots for an evm object

In texmex: Statistical Modelling of Extreme Values

Diagnostic plots for an evm object

Description

Create and display diagnostic plots for an evm object. See plot.evmOpt for further details on what is being plotted.

Usage

## S3 method for class 'evmOpt'
ggplot(
  data,
  mapping,
  which = 1:4,
  main = rep(NULL, 4),
  xlab = rep(NULL, 4),
  nsim = 1000,
  alpha = 0.05,
  jitter.width = 0,
  jitter.height = 0,
  ptcol = "blue",
  span = 2/3,
  col = "light blue",
  fill = "orange",
  plot. = TRUE,
  ncol = 2,
  nrow = 2,
  ...,
  environment
)

Arguments

data	An object of class evm.
mapping, environment	ignored
which	Which plots to produce. Defaults to which=1:4.
main	Main titles. Should have length 4.
xlab	Labels for x-axes.
nsim	Number of simulated datasets to create to form tolerance regions.
alpha	Used to compute coverage of pointwise confidence intervals.
jitter.width, jitter.height	Used to control the amount of jittering of points in the plots of the residuals versus covariates (when covariates are in the model). Defaults to jitter.width=0, jitter.height = 0.
ptcol	Colour for points. Defaults to ptcol="blue".
span	Passed to the loess smoother and defaults to span=2/3. Sometimes this choice is poor: if the loess smoother looks wrong, try span=1.
col	Colour for lines. Defaults to col="light blue".
fill	Colour for confidence regions. Defaults to fill="orange"
plot.	Whether or not to display the output. Defaults to plot.=TRUE. If the display doesn't have the desired row and column layout, the user should specify plot.=FALSE, asign the output to an object, and use grid.arrange to display it.
ncol	The number of columns wanted in the resulting plot. Defaults to ncol=2. This argument is passed into grid.arrange.
nrow	The number of rows wanted in the resulting plot. Defaults to nrow=2. This argument is passed into grid.arrange.
...	Other arguments passed through to underlying plot functions.

Details

The function attempts to arrange the plots nicely. If the output isn't what was wanted, the function returns the graphs to the user as a list so that the user can use grid.arrange directly. Also, if you have one or more covariates in the model and the loess smoother looks wrong, try setting span=1.

See Also

plot.evmOpt

texmex documentation built on June 22, 2024, 12:26 p.m.