segmented (version 0.2-5)

plot.segmented: Plot method for segmented objects

Description

Takes a fitted segmented object returned by segmented() and plots (or add) the fitted broken-line for the selected segmented term.

Usage

## S3 method for class 'segmented':
plot(x, term = NULL, se = FALSE, const = coef(x)["(Intercept)"],
  add = FALSE, linkinv = FALSE, show.gap=TRUE, rev.sgn=FALSE,
  n.points = 10, ...)

Arguments

x

a fitted segmented object

term

the segmented variable whose piece-wise relationship has to be plotted. If there is a single segmented variable, term can be omitted

se

when TRUE pointwise confidence intervals are drawn. Currently unimplemented.

const

constant to add to each fitted segmented relationship (on the scale of the linear predictor) before plotting

add

when TRUE the fitted lines are added to the current device

linkinv

when TRUE, the fitted lines are (possibly) tranformed on the inverse link scale before plotting; in this case it could be useful to increase n.points to get smooth curves

show.gap

should the gap between the fitted lines at the estimated breakpoints to be shown?

rev.sgn

when TRUE it is assumed that current term is `minus' the actual segmented variable, therefore the sign is reversed before plotting. This is useful when a null-constraint has been set on the last slope

n.points

number of points where the fitted lines have to be computed

...

other graphics parameters to pass on to plotting commands

Value

• None.

Details

Produces (or adds to the current device) the fitted segmented relationship between the response and the selected term. If the fitted model includes just a single `segmented' variable, term may be omitted. Due to the parameterization of the segmented terms, sometimes the fitted lines may not appear to join at the estimated breakpoints. If this is the case, the apparent `gap' would indicate some lack-of-fit. However such `gap' may be hidden by setting show.gap=FALSE: in this case the new fitted values are re-computed by means of a linear model fit. Note that, as hidding could lead to a wrong impression of the fit, show.gap=TRUE is suggested only when the gap coefficients are nonsignificant.

See Also

lines.segmented

Examples

set.seed(1234)
z<-runif(100)
y<-rpois(100,exp(2+1.8*pmax(z-.6,0)))
o<-glm(y~z,family=poisson)
o.seg<-segmented(o,seg.Z=~z,psi=list(z=.5))
par(mfrow=c(1,2))
plot(o.seg)
plot(z,y)
plot(o.seg,add=TRUE,linkinv=TRUE,lwd=2,col=2)
lines(o.seg,col=2,pch=19,bottom=FALSE,lwd=2)