inspect {asuR}R Documentation

Plots to inspect fitted models

Description

This functions can be used to validate some assumptions of linear regressions. The idea of this functions is to illustrate possibilities, not to provide fully functional code. The functions work for regression models with continous variables only.

Usage

inspect(mymodel, which = c("select", "sequence", "all"), id = c("all", "none"), ...)
irp(mymodel)
ryp(mymodel, id= c("none", "all"))
ilp(mymodel)
prp(mymodel)

Arguments

mymodel an object of class glm or lm, usually, a result of a call to the function glm or lm
which a character string; one of the following: description{
select
select a diagnostic plot from a menu with all plots available (also repeatedly selecting the same plot is possible)
sequence
plot all available diagnostic plots in a sequence (with the possibility to skip a plot or quit)
all
plot all available diagnositc plots without user interaction (mainly useful for printing, see examples)
}
id a character string or numeric value; in which panel should it be possible to interactively identify values
... further arguments

Details

The function inspect calls other functions (depending on their class) to inspect the model assumptions.

Value

A list with identified values (row names), one slot for each inspection function.

Note

On some devices you are not allowed to resize the window before interactively selecting points

Author(s)

thomas.fabbro@unibas.ch

See Also

plot

Examples


 data(mytrees)
 model <- lm(log(Volume) ~ log(Girth) + log(Height),  data=mytrees)
## inspect(model)

### for printing
##pdf(~/temp/diagnostic.pdf)
##inspect(model)
##dev.off()

[Package asuR version 0.08-24 Index]