xpred.itree: Return Cross-Validated Predictions

Description

Gives the predicted values for an itree fit, under cross validation, for a set of complexity parameter values. Similar to xpred.rpart but will stop and print an error message when given a fit for which cp is not defined. If fit has a penalty, then all the cross-validations use the same penalty and penalization constant found in the fit object.

Usage

xpred.itree(fit, xval=10, cp)

Arguments

fit

a itree object.

xval

number of cross-validation groups. This may also be an explicit list of integers that define the cross-validation groups.

the desired list of complexity values. By default it is taken from the cptable component of the fit.

Value

A matrix with one row for each observation and one column for each complexity value.

Details

From rpart: Complexity penalties are actually ranges, not values. If the cp values found in the table were $.36$, $.28$, and $.13$, for instance, this means that the first row of the table holds for all complexity penalties in the range $[.36, 1]$, the second row for cp in the range $[.28, .36)$ and the third row for $[.13,.28)$. By default, the geometric mean of each interval is used for cross validation.

Examples

Run this code

#rpart's example:
fit <- itree(Mileage ~ Weight, car.test.frame)
xmat <- xpred.itree(fit)
xerr <- (xmat - car.test.frame$Mileage)^2
apply(xerr, 2, sum)   # cross-validated error estimate

# approx same result as rel. error from printcp(fit)
apply(xerr, 2, sum)/var(car.test.frame$Mileage) 
printcp(fit)