ErrorClustering: Get the approximate error of a clustering algorithm

In mgerault/pRolocExtra: New functions to pRoloc package

Get the approximate error of a clustering algorithm

Description

ErrorClustering perform n loop to calculate the loss of the clustering algorithm from the pRoloc package. Take only the protein from known locations, separate randomly in 80/20 train set / test set. Then count the number of mistake made by the algorithm.

Usage

ErrorClustering(data, n, cmet = "knn", train_size = 0.8)

Arguments

data	A MSnSet object
n	An integer which correspond to the number of loop
cmet	A character which correspond to the clusterig method of the pRoloc package knn, svm, naiveBayes, perTurbo, nnet (neural network) and rf (random forest).
train_size	A numeric between 0 and 1 corresponding to train size in percentage

Value

a list containing the loss and the clustering score of the proteins which were miss located at each loop; The mean loss and the mean clustering score of the proteins which were miss located

See Also

svmOptimisation from pRoloc package and datavisupca for more details

Examples

library(pRolocExtra)
err_func(tan2009r1, 5)

mgerault/pRolocExtra documentation built on Sept. 15, 2022, 9:26 a.m.