Data Mining Algorithms In R/Packages/CCMtools/Percent.bad.and.false.classif.per.cluster

Description

edit

This function computes the percentage of bad and false classification of a sequence of clusters (new.cl) according to a reference sequence (cl).

Usage

edit

Percent.bad.and.false.classif.per.cluster(cl, new.cl)

Arguments

edit
  • cl Reference sequence of clusters.
  • new.cl Sequence of clusters to be compared to the reference sequence.

Value

edit

Returns a list containing the following elements:

  • tot Global percentage of bad classification
  • BadPerCluster Percentage of bad classification per cluster
  • FalsePerCluster Percentage of false classification per cluster
  • mat.att Global matrix of attribution (row = cl, colomn = new.cl)

Author

edit

M. Vrac (mathieu.vrac@lsce.ipsl.fr))

See Also

edit
  1. learn.and.project.clusters