plot.profile {stats} R Documentation

Plotting Functions for 'profile' Objects

Description

plot and pairs methods for objects of class "profile".

Usage

## S3 method for class 'profile'
plot(x, ...)
## S3 method for class 'profile'
pairs(x, colours = 2:3, which = names(x), ...)


Arguments

 x an object inheriting from class "profile". colours colours to be used for the mean curves conditional on x and y respectively. which names or number of parameters in pairs plot ... arguments passed to or from other methods.

Details

This is the main plot method for objects created by profile.glm. It can also be called on objects created by profile.nls, but they have a specific method, plot.profile.nls.

The pairs method shows, for each pair of parameters x and y, two curves intersecting at the maximum likelihood estimate, which give the loci of the points at which the tangents to the contours of the bivariate profile likelihood become vertical and horizontal, respectively. In the case of an exactly bivariate normal profile likelihood, these two curves would be straight lines giving the conditional means of y|x and x|y, and the contours would be exactly elliptical. The which argument allows you to select a subset of parameters; the default corresponds to the set of parameters that have been profiled.

Author(s)

Originally, D. M. Bates and W. N. Venables for S (in 1996). Taken from MASS where these functions were re-written by B. D. Ripley for R (by 1998).

profile.glm, profile.nls.

Examples

## see ?profile.glm for another example using glm fits.

## a version of example(profile.nls) from R >= 2.8.0
fm1 <- nls(demand ~ SSasympOrig(Time, A, lrc), data = BOD)
pr1 <- profile(fm1, alphamax = 0.1)
stats:::plot.profile(pr1) ## override dispatch to plot.profile.nls
pairs(pr1) # a little odd since the parameters are highly correlated

## an example from ?nls
x <- -(1:100)/10
y <- 100 + 10 * exp(x / 2) + rnorm(x)/10
nlmod <- nls(y ~  Const + A * exp(B * x), start=list(Const=100, A=10, B=1))
pairs(profile(nlmod))

## example from Dobson (1990) (see ?glm)
counts <- c(18,17,15,20,10,20,25,13,12)
outcome <- gl(3,1,9)
treatment <- gl(3,3)
## this example is only formally a Poisson model. It is really a
## comparison of 3 multinomials. Only the interaction parameters are of
## interest.
glm.D93i <- glm(counts ~ outcome * treatment, family = poisson())
pr1 <- profile(glm.D93i)
pr2 <- profile(glm.D93i, which=6:9)
plot(pr1)
plot(pr2)
pairs(pr1)
pairs(pr2)


[Package stats version 4.4.0 Index]