[R] errorest

From: Kuhn, Max <Max.Kuhn_at_pfizer.com>
Date: Fri 24 Jun 2005 - 01:17:58 EST


Jenny,

"It didn't work" and "They worked" aren't very specific. Also, the package name is ipred and the function is errorest.

The estimator entry on the man page for errorest has:

   'cv' cross-validation, 'boot' bootstrap or '632plus' bias corrected bootstrap (classification only).

Note the *or*. I tried the analysis of the iris data from the man page with your estimator specification:

> testing <- errorest(Species ~ ., data=iris, model=lda, + estimator = c("boot","632plus"), predict= mypredict.lda) > testing

Call:
errorest.data.frame(formula = Species ~ ., data = iris, model = lda,

    predict = mypredict.lda, estimator = c("boot", "632plus"))

         Bootstrap estimator of misclassification error 
         with 25 bootstrap replications

Misclassification error: 0.0235
Standard deviation: 0.0028

Call:
errorest.data.frame(formula = Species ~ ., data = iris, model = lda,

    predict = mypredict.lda, estimator = c("boot", "632plus"))

         .632+ Bootstrap estimator of misclassification error 
         with 25 bootstrap replications

Misclassification error:  0.0222 
>               

> unclass(testing)
$boot
$boot$error
[1] 0.02351852

$boot$sd
[1] 0.002847447

$boot$bc632plus
[1] FALSE
$boot$nboot
[1] 25

$"632plus"
$"632plus"$error
[1] 0.02222817

$"632plus"$nboot
[1] 25

$"632plus"$bc632plus
[1] TRUE
$call
errorest.data.frame(formula = Species ~ ., data = iris, model = lda,

    predict = mypredict.lda, estimator = c("boot", "632plus"))

Is this consistent with your results?

Max

LEGAL NOTICE\ Unless expressly stated otherwise, this messag...{{dropped}}



R-help@stat.math.ethz.ch mailing list
https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html Received on Fri Jun 24 01:31:06 2005

This archive was generated by hypermail 2.1.8 : Fri 03 Mar 2006 - 03:33:00 EST