Cluster {MVar}

Cluster Analysis.

Description

Performs hierarchical and non-hierarchical cluster analysis in a data set.

Usage

Cluster(data, titles = NA, hierarquic = TRUE, analysis = "Obs",  
        cor.abs = FALSE, normalize = FALSE, distance = "euclidean",  
        method = "complete", horizontal = FALSE, num.groups = 0,
        lambda = 2, savptc = FALSE, width = 3236, height = 2000, 
        res = 300, casc = TRUE)

Arguments

data

Data to be analyzed.

titles

Titles of the graphics, if not set, assumes the default text.

hierarquic

Hierarchical groupings (default = TRUE), for non-hierarchical groupings (method K-Means), only for case 'analysis' = "Obs".

analysis

"Obs" for analysis on observations (default), "Var" for analysis on variables.

cor.abs

Matrix of absolute correlation case 'analysis' = "Var" (default = FALSE).

normalize

Normalize the data only for case 'analysis' = "Obs" (default = FALSE).

distance

Metric of the distances in case of hierarchical groupings: "euclidean" (default), "maximum", "manhattan", "canberra", "binary" or "minkowski". Case Analysis = "Var" the metric will be the correlation matrix, according to cor.abs.

method

Method for analyzing hierarchical groupings: "complete" (default), "ward.D", "ward.D2", "single", "average", "mcquitty", "median" or "centroid".

horizontal

Horizontal dendrogram (default = FALSE).

num.groups

Number of groups to be formed.

lambda

Value used in the minkowski distance.

savptc

Saves graphics images to files (default = FALSE).

width

Graphics images width when savptc = TRUE (defaul = 3236).

height

Graphics images height when savptc = TRUE (default = 2000).

res

Nominal resolution in ppi of the graphics images when savptc = TRUE (default = 300).

casc

Cascade effect in the presentation of the graphics (default = TRUE).

Value

Several graphics.

tab.res

Table with similarities and distances of the groups formed.

groups

Original data with groups formed.

res.groups

Results of the groups formed.

R.sqt

Result of the R squared.

sum.sqt

Total sum of squares.

mtx.dist

Matrix of the distances.

Author(s)

Paulo Cesar Ossani

References

Rencher, A. C. Methods of multivariate analysis. 2th. ed. New York: J.Wiley, 2002. 708 p.

Mingoti, S. A. analysis de dados atraves de metodos de estatistica multivariada: uma abordagem aplicada. Belo Horizonte: UFMG, 2005. 297 p.

Ferreira, D. F. Estatistica Multivariada. 2a ed. revisada e ampliada. Lavras: Editora UFLA, 2011. 676 p.

Examples

data(DataQuan) # set of quantitative data

data <- DataQuan[,2:8]

rownames(data) <- DataQuan[1:nrow(DataQuan),1]

res <- Cluster(data, titles = NA, hierarquic = TRUE, analysis = "Obs",
               cor.abs = FALSE, normalize = FALSE, distance = "euclidean", 
               method = "ward.D", horizontal = FALSE, num.groups = 2,
               savptc = FALSE, width = 3236, height = 2000, res = 300, 
               casc = FALSE)

print("R squared:"); res$R.sqt
# print("Total sum of squares:"); res$sum.sqt
print("Groups formed:"); res$groups
# print("Table with similarities and distances:"); res$tab.res
# print("Table with the results of the groups:"); res$res.groups
# print("Distance Matrix:"); res$mtx.dist 
 
write.table(file=file.path(tempdir(),"SimilarityTable.csv"), res$tab.res, sep=";",
            dec=",",row.names = FALSE) 
write.table(file=file.path(tempdir(),"GroupData.csv"), res$groups, sep=";",
            dec=",",row.names = TRUE) 
write.table(file=file.path(tempdir(),"GroupResults.csv"), res$res.groups, sep=";",
            dec=",",row.names = TRUE)