dudi.acm {ade4}R Documentation

Multiple Correspondence Analysis

Description

dudi.acm performs the multiple correspondence analysis of a factor table.
acm.burt an utility giving the crossed Burt table of two factors table.
acm.disjonctif an utility giving the complete disjunctive table of a factor table.
boxplot.acm a graphic utility to interpret axes.

Usage

dudi.acm (df, row.w = rep(1, nrow(df)), scannf = TRUE, nf = 2)
acm.burt (df1, df2, counts = rep(1, nrow(df1))) 
acm.disjonctif (df) 
## S3 method for class 'acm':
boxplot(x, xax = 1, ...) 

Arguments

df, df1, df2 data frames containing only factors
row.w, counts vector of row weights, by default, uniform weighting
scannf a logical value indicating whether the eigenvalues bar plot should be displayed
nf if scannf FALSE, an integer indicating the number of kept axes
x an object of class acm
xax the number of factor to display
... further arguments passed to or from other methods

Value

dudi.acm returns a list of class acm and dudi (see dudi) containing
cr a data frame which rows are the variables, columns are the kept scores and the values are the correlation ratios

Author(s)

Daniel Chessel
Anne B Dufour dufour@biomserv.univ-lyon1.fr

References

Tenenhaus, M. & Young, F.W. (1985) An analysis and synthesis of multiple correspondence analysis, optimal scaling, dual scaling, homogeneity analysis ans other methods for quantifying categorical multivariate data. Psychometrika, 50, 1, 91-119.

Lebart, L., A. Morineau, and M. Piron. 1995. Statistique exploratoire multidimensionnelle. Dunod, Paris.

See Also

s.chull, s.class

Examples

data(ours)
summary(ours)
boxplot(dudi.acm(ours, scan = FALSE))
## Not run: 
data(banque)
banque.acm <- dudi.acm(banque, scann = FALSE, nf = 3)
scatter.dudi(banque.acm)

apply(banque.acm$cr, 2, mean)
banque.acm$eig[1:banque.acm$nf] # the same thing
boxplot.acm(banque.acm)

scatter(banque.acm)

s.value(banque.acm$li, banque.acm$li[,3])

bb <- acm.burt(banque, banque)
bbcoa <- dudi.coa(bb, scann = FALSE)
plot(banque.acm$c1[,1], bbcoa$c1[,1])
# mca and coa of Burt table. Lebart & coll. section 1.4

bd <- acm.disjonctif(banque)
bdcoa <- dudi.coa(bd, scann = FALSE)
plot(banque.acm$li[,1], bdcoa$li[,1]) 
# mca and coa of disjonctive table. Lebart & coll. section 1.4
plot(banque.acm$co[,1], dudi.coa(bd, scann = FALSE)$co[,1]) 

## End(Not run)

Worked out examples


> library(ade4)
> ### Name: dudi.acm
> ### Title: Multiple Correspondence Analysis
> ### Aliases: dudi.acm acm.burt acm.disjonctif boxplot.acm
> ### Keywords: multivariate
> 
> ### ** Examples
> 
> data(ours)
> summary(ours)
 altit  deniv  cloiso domain boise  hetra  favor  inexp  citat  depart 
 1: 8   1:13   1:12   1: 9   1:10   1:19   1:15   1:20   1:22   AHP:5  
 2:17   2:14   2: 4   2:13   2:15   2: 5   2:12   2:10   2: 7   AM :4  
 3:13   3:11   3:22   3:16   3:13   3:14   3:11   3: 8   3: 4   D  :5  
                                                         4: 5   HP :8  
                                                                HS :4  
                                                                I  :5  
                                                                S  :7  
> boxplot(dudi.acm(ours, scan = FALSE))
> 
> data(banque)
> banque.acm <- dudi.acm(banque, scann = FALSE, nf = 3)
> scatter.dudi(banque.acm)
> 
> apply(banque.acm$cr, 2, mean)
       RS1        RS2        RS3 
0.17346599 0.11838319 0.09825814 
> banque.acm$eig[1:banque.acm$nf] # the same thing
[1] 0.17346599 0.11838319 0.09825814
> boxplot.acm(banque.acm)
> 
> scatter(banque.acm)
> 
> s.value(banque.acm$li, banque.acm$li[,3])
> 
> bb <- acm.burt(banque, banque)
> bbcoa <- dudi.coa(bb, scann = FALSE)
> plot(banque.acm$c1[,1], bbcoa$c1[,1])
> # mca and coa of Burt table. Lebart & coll. section 1.4
> 
> bd <- acm.disjonctif(banque)
> bdcoa <- dudi.coa(bd, scann = FALSE)
> plot(banque.acm$li[,1], bdcoa$li[,1]) 
> # mca and coa of disjonctive table. Lebart & coll. section 1.4
> plot(banque.acm$co[,1], dudi.coa(bd, scann = FALSE)$co[,1]) 
> 
> 
> 
> 

[Package ade4 Index]