plotROC: Perform biomarker analysis and plot ROC curve

In wenbostar/metaX: A flexible and comprehensive software for processing omics data.

Perform biomarker analysis and plot ROC curve

Description

Perform biomarker analysis and plot ROC curve

Usage

plotROC(para, group, fID, valueID = "valueNorm", method = "rf",
  ratio = 2/3, k = 100, plotCICurve = TRUE, ...)

Arguments

para	An object of metaXpara
group	The sample class used
valueID	The column name used
method	Method for modeling
ratio	(The number of sample for training)/(The number of sample for Testing)
k	The number of sampling. Default is 100.
plotCICurve	Whether plot confidence interval curve in the ROC figure
...	Additional parameters

Value

The result of feature selection and modeling

Examples

para <- new("metaXpara")
pfile <- system.file("extdata/MTBLS79.txt",package = "metaX")
sfile <- system.file("extdata/MTBLS79_sampleList.txt",package = "metaX")
rawPeaks(para) <- read.delim(pfile,check.names = FALSE)
sampleListFile(para) <- sfile
para <- reSetPeaksData(para)
para <- missingValueImpute(para)
para <- transformation(para,valueID = "value")
# library(doMC)
# registerDoMC(cores = 4)
res <- featureSelection(para,group=c("C","S"), plot_roc=FALSE)
plotROC(para,group=c("C","S"),valueID="value",fID=res$optVariables)

wenbostar/metaX documentation built on July 4, 2023, 7:50 p.m.