Data Mining Algorithms In R/Packages/CCMtools/Info.Criterion

Description

edit

This function computes the Information Criterion (IC) of a clustering result.

Usage

edit

Info.Criterion(NS, DataS, r, totCL, Nc, cl)

Arguments

edit
  • NS Number of locations (i.e., weather stations) for the local-scale time series on which IC is calculated.
  • DataS Dataset corresponding to local-scale (station) data on which IC is calculated. This is a matrix NS*NN, where NN is the number of days (i.e., length of the time series).
  • r Value for which the IC is calculated (see details).
  • totCL Vector of numbers of elements (e.g., days) in each cluster.
  • Nc Number of clusters.
  • cl Vector containing the sequence of clusters (length(cl) is NN).

Details

edit
    The IC is computed as $IC = Sum_i=1^K |n_i,r - (p_r * n_i)|$, where
    $n_i,r$ = \# of days in cluster $i$ that receive a rainfall amount > r
    $p_r$ = proba of such rainy days in the whole population
    $n_i$ = \# of days in cluster i

Author

edit

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

See Also

edit
  1. CCM