| divisiveHC.details {LearnClust} | R Documentation |
To explain the divisive hierarchical clusterization algorithm by distance and approach.
Description
To explain the complete divisive hierarchical clusterization algorithm by choosing distance and approach types.
Usage
divisiveHC.details(data, distance, approach)
Arguments
data |
could be a numeric vector, a matrix or a numeric data frame. It will be transformed into matrix and list to be used. |
distance |
is a string. It chooses the distance to use. |
approach |
is a string. It chooses the approach to use. |
Details
This function is the main part of the divisive hierarchical clusterization method. It explains the theoretical algorithm step by step.
1 - The function transforms data in useful object to be used.
2 - It creates a cluster that includes every simple elements.
3 - It initializes posible clusters using the initial elements.
4 - It calculates a matrix distance with the clusters created in the 3rd step.
5 - It chooses the maximal distance value and gets the clusters to be divided.
6 - It divides the cluster into two new complementary clusters and updates the clusters list.
6 - It repeats these steps until every cluster can't be divided again. The solution includes every simple cluster.
Value
A list with the divided clusters. Explanation
Author(s)
Roberto Alcántara roberto.alcantara@edu.uah.es
Juan José Cuadrado jjcg@uah.es
Universidad de Alcalá de Henares
Examples
a <- c(1,2,1,3,1,4,1,5,1,6)
matrixA <- matrix(a,ncol=2)
dataFrameA <- data.frame(matrixA)
divisiveHC.details(a,'EUC','MAX')
divisiveHC.details(matrixA,'MAN','AVG')
divisiveHC.details(dataFrameA,'CHE','MIN')