Produce a shaded Curve

Sometimes the area under a density curve has to be color shaded, for instance to illustrate a p-value or a specific region under the normal curve. This function draws a curve corresponding to a function over the interval [from, to]. It can plot also an expression in the variable xname, default x.

Usage

shade(expr, col = par("fg"), breaks, density = 10, n = 101, xname = "x", ...)

Arguments

expr

the name of a function, or a call() or an expression() written as a function of x which will evaluate to an object of the same length as x.

col

color to fill or shade the shape with. The default is taken from par("fg").

breaks

numeric, a vector giving the breakpoints between the distinct areas to be shaded differently. Should be finite as there are no plots with infinite limits.

density

the density of the lines as needed in polygon.

n

integer; the number of x values at which to evaluate. Default is 101.

xname

character string giving the name to be used for the x axis.

...

the dots are passed on to polygon().

Value

A list with components x and y of the points that were drawn is returned invisibly.

Details

Useful for shading the area under a curve as often needed for explaining significance tests.

See also

polygon(), curve()

Other plot.distribution: plotFun(), plotProbDist()

Examples

curve(dt(x, df=5), xlim=c(-6,6),
      main=paste("Student t-Distribution Probability Density Function, df = ", 5, ")", sep=""),
      type="n", las=1, ylab="probability", xlab="t")

shade(dt(x, df=5), breaks=c(-6, qt(0.025, df=5), qt(0.975, df=5), 6),
      col=c("deeppink4", "skyblue3"), density=c(20, 7))