lmomco (version 2.3.1)

add.lmomco.axis: Add an lmomco Axis to a Plot

Description

This function provides special support for adding probability-like axes to an existing plot. The function supports a recurrence interval (RI) axis, normal probability axis (NPP), and standard normal variate (SNV) axis. The function is built around the interface model that standard normal transformation of the values for the respective axis controlled by this function are being plotted; this means that qnorm() should be wrapped on the values of nonexceedance probability. This is an ease oversight to make (see Examples section below and note use of qnorm(pp)).

The function provides a convenient interface for labeling and titling two axes, so adjustments to default margins might be desired. The pertinent control is achieved using the par() function, which might be of the form par(mgp=c(3,0.5,0), mar=c(5,4,4,3)) say for plotting the lmomco axis both on the left and right (see z.par2cdf for an example).

Usage

add.lmomco.axis(side=1, twoside=FALSE, side.type=c("NPP", "RI", "SNV"),
                otherside.type=c("NA", "RI", "SNV", "NPP"),
                alt.lab=NA, alt.other.lab=NA, npp.as.aep=FALSE,
                NPP.control=NULL, RI.control=NULL, SNV.control=NULL, ...)

Arguments

side

The side of the plot (1=bottom, 2=left, 3=top, 4=right).

twoside

A logical triggering whether the tick marks are echoed on the opposite side. This value is forced to FALSE if otherside.type is not "NA".

side.type

The axis type for the primary side.

otherside.type

The optional axis type for the opposite side. The default is a literal not applicable.

alt.lab

A short-cut to change the axis label without having to specify a *.control argument and its label attribute. The label attribute of alt.lab is not NA is used instead of the defaults. This argument overrides behavior of the otherside.type labeling so use of alt.lab only makes sense if otherside.type is left as NA.

alt.other.lab

Similar to alt.lab but can house an alternative label (see Examples.

npp.as.aep

Convert nonexceedance probability to exceedance probability, which is a que for alt.other.lab and nonexceedance probabilities are changed by \(1-F\), but the real coordinates for plotting remain in the nonexceedance probability context.

NPP.control

An optional R list used to influence the NPP axis.

RI.control

An optional R list used to influence the RI axis.

SNV.control

An optional R list used to influence the SNV axis.

...

Additional arguments that are passed to the R function Axis.

Value

No value is returned. This function is used for its side effects.

See Also

prob2T, T2prob, add.log.axis

Examples

# NOT RUN {
par(mgp=c(3,0.5,0)) # going to tick to the inside, change some parameters
X <- sort(rnorm(65)); pp <- pp(X) # generate synthetic data
plot(qnorm(pp), X, xaxt="n", xlab="", ylab="QUANTILE", xlim=c(-2,3))
add.lmomco.axis(las=2, tcl=0.5, side.type="RI", otherside.type="NPP")
par(mgp=c(3,1,0)) # restore defaults

# }
# NOT RUN {
opts <- options(scipen=6); par(mgp=c(3,0.5,0))
X <- sort(rexp(65, rate=.0001))*100; pp <- pp(X) # generate synthetic data
plot(qnorm(pp), X, yaxt="n", xaxt="n", xlab="", ylab="", log="y")
add.log.axis(side=2,    tcl=+0.8*abs(par()$tcl),         two.sided=TRUE)
add.log.axis(logs=c(1), tcl=-0.5*abs(par()$tcl), side=2, two.sided=TRUE)
add.log.axis(logs=c(1), tcl=+1.3*abs(par()$tcl), side=2, two.sided=TRUE)
add.log.axis(logs=1:8, side=2, make.labs=TRUE, las=1, label="QUANTILE")
add.lmomco.axis(las=2, tcl=0.5, side.type="NPP", npp.as.aep=TRUE)
options(opts)
par(mgp=c(3,1,0)) # restore defaults
# }