sim2dis {optiSel}

Converts a Similarity Matrix into a Dissimilarity Matrix

Description

Converts a similarity matrix (e.g. a kinship matrix) into a dissimilarity matrix.

Usage

sim2dis(f, a=4.0, baseF=0.03, method=1)

Arguments

f

Similarity matrix.

a

Exponent

baseF

Old inbreeding not measured by f

method

Either 1 or 2

Details

This function converts a similarity matrix f with values between 0 and 1 (e.g. a kinship matrix) into a dissimilarity matrix. At first, the similarity is adjusted as

f <- baseF + (1-baseF)*f.

Then, for Method 1, the dissimilarity between individuals i and j is computed as

Dij = (-log(fij))^a,

whereas for Method 2, the dissimilarity is computed as

Dij = sqrt((fii+fjj)/2-fij)^a.

Although Method 2 may provide lower stress values in some cases, Method 1 has the advantage that the area reflects the diversity of a population more reasonable.

Value

Dissimilarity matrix D.

Author(s)

Robin Wellmann

Examples

data(map)
dir   <- system.file("extdata", package = "optiSel")
files <- file.path(dir, paste("Chr", unique(map$Chr), ".phased", sep=""))
f     <- segIBD(files, map, minSNP=15, minL=1.0)
D     <- sim2dis(f, 4)

## Multidimensional scaling of animals:

data(Cattle)
library("smacof")
color <- c(Angler="red", Rotbunt="green", Fleckvieh="blue", Holstein="black")
col   <- color[as.character(Cattle$Breed)]
Res   <- smacofSym(D, itmax = 5000, eps = 1e-08)
plot(Res$conf, pch=18, col=col, main="Multidimensional Scaling", cex=0.5)
mtext(paste("segIBD Stress1 = ", round(Res$stress,3)))