mratios (version 1.4.4)

plot.sci.ratio: Plot output for sci.ratio and sci.ratio.gen

Description

Plot the intervals returned by sci.ratio

Usage

# S3 method for sci.ratio
plot(x, rho0 = 1, rho0lty=2, rho0lwd=1, rho0col="black",
 CIvert = FALSE, CIlty = 1, CIlwd = 1, CIcex = 1, CIpch=16,
 main = NULL, ylab = NULL, xlab = NULL, sub = NULL, length=NULL,
 sortby=NULL, decreasing=NULL, ...)

Value

A plot of the confidence intervals in the sci.ratio object.

Arguments

x

an object of class "sci.ratio" as can be obtained by calling the function sci.ratio

rho0

a single numeric value or vector of values defining the hypothesized ratio

rho0lty

integer values to specify the line type for the rho0 line(s)

rho0lwd

integer values to specify the line width for the rho0 line(s)

rho0col

character vector to specify the colour for the rho0 line(s)

CIvert

logical, CI are plotted horizontal if CIvert=FALSE and vertical otherwise

CIlty

numeric value, giving the line type of the plotted confidence interval, see argument lty in ?par

CIlwd

numeric value, giving the line width of the plotted confidence interval, see argument lwd in ?par

CIcex

a single numeric value: by which amount the symbols in the CI shall be scaled relative to the default (see argument cex in ?par)

CIpch

the symbol to be used for the point estimate, see pch in ?points

main

character string to be plotted as main title of the plot

ylab

character string, label of the y axis (ignored if CIvert=TRUE)

xlab

character string, label of the x axis (ignored if CIvert=FALSE)

sub

as in plot

length

a numeric value, specifying the length/2 of the bars at the ends of the confidence intervals in inches

sortby

a character string, one of "estimate", "lower, or "upper"; if specified, the results are ordered by magnitude of estimates, lower or upper limits

decreasing

logical, to be passed to order, if sortby is specified, ignored otherwise

...

further arguments to be passed to axis()

Author

Frank Schaarschmidt

Details

Too long names of the contrasts/comparisons should be avoided, otherwise use par() to change plot parameters.

References

plot.hmtest(multcomp)

Examples

library(mratios)

data(angina)

aCI<-sci.ratio(response~dose, data=angina, type="Dunnett",
 alternative="greater")

# Visualize testing for superiority

plot(aCI, rho0=1.25, rho0lty=3)