agricolae (version 1.0-1)

coeff.diana: Selection of the best method for the dissimilarity distance calculation

Description

A method based on the Divisive Coefficient by the DIANA method. It uses the methods of distance provided by function DIST of library CLUSTER. Also, it reports the first selected distance based on highest value of the coefficient.

Usage

coeff.diana(data)

Arguments

data

data.frame

Value

• data
  each column represents a variable

Details

distance: euclidean, maximum, manhattan, canberra, binary, minkowski.

References

Finding Groups in Data. An Introduction to Cluster Analysis. Leonard Kaufaman. 1990, Jhon Wiley ans Sons, Inc.

Examples

library(cluster)
library(agricolae)
# Example 1
data.1 <- rbind(c(1,1,1,1,1),c(1,1,0,1,1),c(0,0,1,0,0),c(1,0,1,0,1),
c(0,0,0,0,0),c(1,1,1,1,1))
dimnames(data.1)<-list(c("A","B","C","D","E","F"),c("m1","m2","m3","m4","m5"))
distance<-coeff.diana(data.1)
h<-hclust(distance,method="mcquitty")
#startgraph
plot(h)
#endgraph

# Example 2
data.2 <- rbind(c(1200,2,6,20),c(1500,1,3,10),c(1400,1,3,20),c(1600,2,5,15),
c(1000,1,2,5))
dimnames(data.2)<-list(c("A","B","C","D","E"),c("v1","v2","v3","v4"))
distance<-coeff.diana(data.2)

# Example 3
data.3 <-rbind(c(4,2,6,15),c(NA,1,3,18),c(2,1,NA,16),c(6,2,5,15),c(NA,1,2,NA),
c(4,2,6,10))
dimnames(data.3)<-list(c("+","+","+",".",".","."),c("x1","x2","x3","x4"))
distance<-coeff.diana(data.3)