weightedHammingDist {compdiagTools}R Documentation

Distance to compare two splits

Description

This function computes a distance between two bipartitions of the same data computed by the ISIS method.

Usage

weightedHammingDist(ires, score.weight = 3)

Arguments

ires Result returned by isis, a binary matrix with one row per split and one column per data point. The value at postion i, j in this matrix associates the sample i to cluster ires[i,j] in split i. In addition, the last column holds DLD scores for each split.
score.weight The weights of splits are used to scale Hamming distances. Splits weights are exponentiated with this value for weighing a distance.

Value

Returns a distance matrix containing distances between samples according to their occurrence in splits.

Note

integrated into compdiagTools by Claudio Lottaz

Author(s)

Joern Toedling

See Also

isis

Examples

# load data select only few genes
library(ALL)
data(ALL)
samples <- 63:128
vars <- apply(exprs(ALL), 1, var)
e <- exprs(ALL)[rank(vars)>12525, samples]

# determine isis splits and compute weights accordingly
library(isis)
splits <- isis(e)
w <- weightedHammingDist(splits)

[Package compdiagTools version 1.5.3 Index]