MCRestimate (version 2.26.0)

MCRconfusion: Summary tables for MCRestimate objects

Description

MCRwrongsamples returns a matrix with all the samples that have a higher frequency of being predicted as a member of a wrong class than of the correct class for at least one classification method. MCRconfusion summarizes the result of the vote matrices

Usage

MCRwrongsamples(x, col.names=names(x), rownames.from.object=TRUE, subgroup=NULL, freq=FALSE)
MCRconfusion(x, col.names=names(x), row.names=NULL)

Arguments

x
List of objects of S3 class MCRestimate
col.names
Vector of strings used for column names. The length must match the number of objects in x
rownames.from.object
Logical. If TRUE then the sample names of the MCRestimate object in x are used as row names
subgroup
Logical. If TRUE then only the samples which belongs to the specified group are listed in the table
freq
Logical. If TRUE then the frequency with which each sample in the table has been misclassified will be printed.
row.names
Vector of strings used for row names. If not specified the names of the groups are used

Value

MCRwrongsamples returns a matrix and MCRconfusion returns a confusion matrix.

See Also

MCRestimate

Examples

Run this code
library(golubEsets)
data(Golub_Train)
exSet <- Golub_Train[1:500,]
result1 <- MCRestimate(exSet,"ALL.AML",classification.fun="RF.wrap",cross.outer=3,cross.repeat=2)
result2 <- MCRestimate(exSet,"ALL.AML",classification.fun="PAM.wrap",poss.parameters=list(threshold=c(0.5,1)),cross.inner=3,cross.outer=3,cross.repeat=2)
MCRwrongsamples(list(result1,result2),subgroup="AML",col.names=c("Random Forest","PAM"))
MCRconfusion(list(result1,result2),col.names=c("Random Forest","PAM"))

Run the code above in your browser using DataCamp Workspace