HH: HH_regsubsets – R documentation

Become an expert in R — Interactive courses, Cheat Sheets, certificates and more!

HH_regsubsets

Display tabular results for Best Subsets Regression.

Description

Print a tabular display of the results of Best Subsets Regression. This is an alternate display for the object from the regsubsets function. This function is based on regsubsets. The functions described here are designed for the HH package in R and use the leaps package in R. The leaps package is not in S-Plus, hence these functions do not work in the HH package for S-Plus.

Usage

`summaryHH`(object, ...)

## S3 method for class 'regsubsets'
summaryHH(object,
           names = abbreviate(dimnames(incidence)[[2]], minlength = abbrev),
           abbrev = 1, min.size = 1, max.size = dim(sumry$which)[2],
           statistic = c("bic", "cp", "adjr2", "rsq", "rss", "stderr"),
           las = par("las"),
           cex.subsets = 1, ..., main=statistic)

## S3 method for class 'summaryHH.regsubsets'
plot(x, ...,
           statistic="adjr2", legend=FALSE,
           col="darkgray", cex=1, pch=16,
           col.text="black", cex.text=1, col.abline="darkgray")

Arguments

`object`	An object of class `"regsubsets"`.
`x`	An object of class `"summaryHH.regsubsets"`.
`statistic`	Name of statistic to be plotted for each model.
`...`	Other arguments to be passed down to `subsets.regsubsets` and `plot`.
`names`	Abbreviations of variable names.
`abbrev`	minimum number of letters in each abbreviation.
`min.size`	minimum size subset to plot; default is 1.
`max.size`	maximum size subset to plot; default is number of predictors.
`legend`	logical variable, `TRUE` if the legend should be printed. If the legend is printed, the execution halts until the user clicks an empty space in the graph where the legend should be placed.
`las`	Orientation for model names on graph.
`cex.subsets`	can be used to change the relative size of the characters used to plot the regression subsets; default is `1`.
`main`	`"main"` title for graph.
`col, cex, pch`	`par` values for dot locating statistic.
`col.text, cex.text`	`par` values for abbreviations of models on plot.
`col.abline`	`par` parameters for `abline` when the statistic is `cp`.

Value

summaryHH produces a table of models, with p, rsq, rss, adjr2, cp, bic, stderr for each. plot.summaryHH.regsubsets plots the specified statistic from the summary. All the others are support functions.

Author(s)

Richard M. Heiberger <rmh@temple.edu>