crossval: Cross validation estimation of the misclassification error

In dprep: Data Pre-Processing and Visualization Functions for Classification

Description

Computation of the misclassification error for the LDA, KNN and rpart classifiers by cross validation

Usage

crossval(data, nparts = 10, method = c("lda", "knn", "rpart","logistic","naiveBayes"), 
kvec = 5,maxwts=2500, repet)

Arguments

data	The name of the dataset
nparts	The number of folds in which the dataset is divided. By default nparts=10.
method	The name of the classifier to be used: LDA,KNN, Rpart.
kvec	The number of nearest neighbors to be used for the KNN classifier.
maxwts	The maximum number of iteratios to be used in th computation of the logistic regression
repet	The number of repetitions

Value

Returns the mean misclassification crossvalidation error of the classifier obtained on a given number of repetitions

Author(s)

Edgar Acuna

See Also

cv10log, cv10mlp

Examples

#------10-fold crossvalidation error using the LDA classifier---
data(bupa)
crossval(bupa,method="lda",repet=10)
## Not run: #------5-fold crossvalidation error using the knn classifier---
data(colon)
crossval(colon,nparts=5,method="knn",kvec=3,repet=5)

## End(Not run)

Example output

Warning messages:
1: In rgl.init(initValue, onlyNULL) : RGL: unable to open X11 display
2: 'rgl_init' failed, running with rgl.useNULL = TRUE 
3: .onUnload failed in unloadNamespace() for 'rgl', details:
  call: fun(...)
  error: object 'rgl_quit' not found 
The error estimation in each repetition are:
 [1] 0.3333333 0.3304348 0.3362319 0.3159420 0.3043478 0.3072464 0.3362319
 [8] 0.3159420 0.3101449 0.3130435
The mean error estimation by cross-validation using all the repetititons is: 
[1] 0.3202899
The error estimation in each repetition are:
[1] 0.1935484 0.1451613 0.1774194 0.1451613 0.1774194
The mean error estimation by cross-validation using all the repetititons is: 
[1] 0.1677419

dprep documentation built on May 29, 2017, 11:01 a.m.