R Graphical Manual

Browse All

Last data update: 2014.03.03

R: Mahalanobis distances for grouped data

pairwise.mahalanobis

R Documentation

Mahalanobis distances for grouped data

Description

Returns a square matrix of Mahalanobis distances by doing a pairwise comparison of group means using the correlation between variables.

Usage

pairwise.mahalanobis(x, grouping = NULL, cov = NULL, inverted = FALSE, digits = 5, ...)

Arguments

`x`	vector or matrix of data with N observations and D variables. If grouping is not specified, the first column is used for grouping observations.
`grouping`	vector of characters or values designating group classification for observations.
`cov`	Covariance matrix (DxD) of the distribution
`inverted`	logical. If TRUE, cov is the inverse of the covariance matrix.
`digits`	number of decimals to keep for the means, cov and distance values
`...`	passed to mahalanobis for computing the inverse of the covariance matrix (if inverted is false).

Details

To determine the distance between group i and group j, the difference of group means for each variable are compared. For a (NxD) data matrix with m groups, a matrix of mxD means and a correlation matrix of DxD values are calculated. pairwise.mahalanobis calculates the mahalanobis distance for all possible group combinations and results in a mxm square distance matrix with m choose 2 distinct pairwise measures.

Value

`means`	(mxD) matrix of group means for each variable
`cov`	(DxD) covariance matrix of centered and scaled data, so it's actually the correlation matrix
`distance`	(mxm) matrix of squared mahalanobis distances

Author(s)

Lisa McFerrin

Examples


data(bHLH288)
grouping = t(bHLH288[,1])
bHLH_Seq = as.vector(bHLH288[,2])
bHLH_pah = FactorTransform(bHLH_Seq, alignment=TRUE)

Mahala1 = pairwise.mahalanobis(bHLH_pah, grouping, digits = 3)
D = sqrt(Mahala1$distance)
D

Results


R version 3.3.1 (2016-06-21) -- "Bug in Your Hair"
Copyright (C) 2016 The R Foundation for Statistical Computing
Platform: x86_64-pc-linux-gnu (64-bit)

R is free software and comes with ABSOLUTELY NO WARRANTY.
You are welcome to redistribute it under certain conditions.
Type 'license()' or 'licence()' for distribution details.

R is a collaborative project with many contributors.
Type 'contributors()' for more information and
'citation()' on how to cite R or R packages in publications.

Type 'demo()' for some demos, 'help()' for on-line help, or
'help.start()' for an HTML browser interface to help.
Type 'q()' to quit R.

> library(HDMD)
Loading required package: psych
Loading required package: MASS
> png(filename="/home/ddbj/snapshot/RGM3/R_CC/result/HDMD/pairwise.mahalanobis.Rd_%03d_medium.png", width=480, height=480)
> ### Name: pairwise.mahalanobis
> ### Title: Mahalanobis distances for grouped data
> ### Aliases: pairwise.mahalanobis
> 
> ### ** Examples
> 
> 
> data(bHLH288)
Warning message:
In data(bHLH288) : data set 'bHLH288' not found
> grouping = t(bHLH288[,1])
> bHLH_Seq = as.vector(bHLH288[,2])
> bHLH_pah = FactorTransform(bHLH_Seq, alignment=TRUE)
> 
> Mahala1 = pairwise.mahalanobis(bHLH_pah, grouping, digits = 3)
> D = sqrt(Mahala1$distance)
> D
         [,1]     [,2]     [,3]     [,4]     [,5]
[1,] 0.000000 4.142149 5.077082 5.619281 4.818219
[2,] 4.142149 0.000000 4.300668 5.662378 4.004002
[3,] 5.077082 4.300668 0.000000 6.356415 4.781217
[4,] 5.619281 5.662378 6.356415 0.000000 6.929369
[5,] 4.818219 4.004002 4.781217 6.929369 0.000000
> 
> 
> 
> 
> 
> 
> dev.off()
null device 
          1 
>