Description Usage Arguments Details Value Author(s) References See Also Examples
This function evaluates classifiers built using microarray data and/or clinical predictors, based on several pairs of learning and test data sets.
1 2 |
x |
A n x p matrix giving the gene expression levels of p genes (columns) for n patients (rows). |
y |
A numeric vector of length n giving the class membership of the n patients, coded as 0,...,K-1 (where K is the number of classes). |
z |
A n x q data frame giving the q clinical predictors for the n patients. Nominal variables should be given as factors, variables with an at least ordinal scale should be given as numeric. |
learningsets |
A matrix with niter rows giving the indices of the
arrays to be included in the learning sets for the niter iterations, as generated by the function |
classifier |
The function used to construct a classifier. The function must have the same structure as |
ncomp |
The candidate numbers of PLS components (if PLS dimension reduction is used). |
varsel |
A niter x p matrix giving the indices of the genes ordered by the chosen gene selection criterion. For example, the element in the first row and the first column is the index of the gene that is ranked best using the first learning set. |
nbgene |
The number of genes to use for classifier construction. Default is |
fold |
The number of folds for the pre-validation step. See Boulesteix et al (2008) for more details. Default is |
... |
Other arguments to be passed to the function |
For an overview of different methods used to generate the learning sets defined by generate.learningsets
,
see Boulesteix et al (2007). These methods include (repeated) cross-validation, subsampling, bootstrap sampling.
error |
A numeric vector of length |
bestncomp |
A numeric vector of length |
OOB |
A list of length |
Anne-Laure Boulesteix (http://www.ibe.med.uni-muenchen.de/organisation/mitarbeiter/020_professuren/boulesteix/eng.html)
Boulesteix AL, Porzelius C, Daumer M, 2008. Microarray-based classification and clinical predictors: On combined classifiers and additional predictive value. Bioinformatics 24:1698-1706.
Boulesteix AL, Strobl C, Augustin T, Daumer D, 2008. Evaluating microarray-based classifiers: an overview. Cancer Informatics 6:77-97.
testclass_simul
, simulate
, generate.learningsets
, plsrf_xz_pv
, plsrf_x_pv
, plsrf_xz
, plsrf_x
,
rf_z
, svm_x
, logistic_z
.
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 | # load MAclinical library
# library(MAclinical)
# Generate data
x<-matrix(rnorm(20000),100,200)
z<-matrix(rnorm(500),100,5)
y<-sample(0:1,100,replace=TRUE)
# Generate learningsets (5-fold CV)
my.learningsets<-generate.learningsets(n=100,method="CV",fold=5)
# Evaluate accuracy of the PLS-PV-RF method
my.eval<-testclass(x=x,y=y,z=z,learningsets=my.learningsets,classifier=plsrf_xz_pv,ncomp=5,
varsel=NULL,nbgene=NULL,fold=10)
# With variable selection
my.varsel<-matrix(0,5,200)
for (i in 1:5)
{
my.varsel[i,]<-order(abs(studentt.stat(X=x[my.learningsets[i,],],
L=y[my.learningsets[i,]]+1)),decreasing=TRUE)
}
my.eval<-testclass(x=x,y=y,z=z,learningsets=my.learningsets,classifier=plsrf_xz_pv,ncomp=5,
varsel=my.varsel,nbgene=15,fold=10)
|
Add the following code to your website.
For more information on customizing the embed code, read Embedding Snippets.