varPro (version 1.0.1)

plot.partialpro: Partial Plots for VarPro

Description

Plot partial effects of x-variable(s) from a VarPro analysis.

Usage

# S3 method for partialpro
plot(x, xvar.names, nvar,
  parametric = FALSE, se = TRUE,
  causal = FALSE, subset = NULL, plot.it = TRUE, ...)

Value

No return value, called for the purpose of generating a plot.

Arguments

x

partialpro object returned from a previous call to partialpro.

xvar.names

Names (or integer indices) of the x-variables to plot. Defaults to all variables.

nvar

Number of variables to plot. Defaults to all.

parametric

Logical. Set to TRUE only if the partial effect is believed to follow a polynomial form.

se

Display standard errors?

causal

Display causal estimator?

subset

Optional conditioning factor. Not applicable if parametric = TRUE. May also be a logical or integer vector to subset the analysis.

plot.it

If FALSE, no plot is produced; instead, partial effect values are returned.

...

Additional arguments passed to plot.

Author

Min Lu and Hemant Ishwaran

Details

Generates smoothed partial effect plots for continuous variables. The solid black line represents the estimated partial effect; dashed red lines show an approximate plus-minus standard error band. These standard errors are intended as heuristic guides and should be interpreted cautiously.

Partial effects are estimated nonparametrically using locally fitted polynomial models. This is the default behavior and is recommended when effects are expected to be nonlinear. Use parametric = TRUE if the underlying effect is believed to follow a global polynomial form.

For binary variables, partial effects are shown as boxplots, with whiskers reflecting variability analogous to standard error.

The causal estimator, when requested, displays the baseline-subtracted parametric local effect.

Conditioning is supported via the subset option. When supplied as a factor (with length equal to the original data), the plot is stratified by its levels. Alternatively, subset can be a logical or integer vector indicating the cases to include in the analysis.

References

Ishwaran H. (2025). Multivariate Statistics: Classical Foundations and Modern Machine Learning, CRC (Chapman and Hall), in press.

See Also

partialpro

Examples

# \donttest{
##------------------------------------------------------------------
##
## Boston housing
##
##------------------------------------------------------------------

library(mlbench)
data(BostonHousing)
o.boston <- varpro(medv~.,BostonHousing)
oo.boston <- partialpro(o.boston, nvar=4, learner=rf.learner(o.boston))

oldpar <- par(mfrow=c(2,4))

## parametric local estimation (default)
plot(oo.boston, ylab="parametric est.")

## non-parametric local estimation
plot(oo.boston, parametric=FALSE, ylab="non-parametric est.")

par(oldpar)

##------------------------------------------------------------------
##
## Boston housing with subsetting
##
##------------------------------------------------------------------

library(mlbench)
data(BostonHousing)
o.boston <- varpro(medv~.,BostonHousing)
oo.boston <- partialpro(o.boston, nvar=3, learner=rf.learner(o.boston))

## subset analysis
price <- BostonHousing$medv
pricef <- factor(price>median(price), labels=c("low priced","high priced"))
oldpar <- par(mfrow=c(1,1))
plot(oo.boston, subset=pricef, nvar=1)
par(oldpar)

##------------------------------------------------------------------
##
## veteran data with subsetting using celltype as a factor
##
##------------------------------------------------------------------

data(veteran, package = "randomForestSRC")
dta <- veteran
dta$celltype <- factor(dta$celltype)
o.vet <- varpro(Surv(time, status)~., dta)
oo.vet <- partialpro(o.vet, nvar=6, nsmp=Inf, learner=rf.learner(o.vet))

## partial effects, with subsetting
oldpar <- par(mfrow=c(2,3))
plot(oo.vet, subset=dta$celltype)
par(oldpar)

## causal effects, with subsetting
oldpar <- par(mfrow=c(2,3))
plot(oo.vet, subset=dta$celltype, causal=TRUE)
par(oldpar)

# }