PTAk (version 1.2-6)

COS2-CTR: Interpretation summaries

Description

After a FCA2, a SVDgen, a FCAk or a PTAk computes the traditional guides for interpretations used in PCA and correspondence analysis: COS2 or the percentage of variability rebuilt by the component and CTR or the amount of contribution towards that component.

Usage

COS2(solu, mod=1, solnbs=2:4, FCA=FALSE)
CTR(solu, mod=1, solnbs=1:4)

Arguments

solu

an object inheriting from class PTAk, representing a generalised singular value decomposition

mod

an integer representing the mode number entry, 1 is row, 2 columns, ...

solnbs

a vector of integers representing the tensor numbers in the listing summary

FCA

if TRUE substract the global independence to the data before computing the variability

Value

• a matrix whose columns are the COS2 or CTR as per thousands (‰) for the mode considered

encoding

utf-8

Details

Classical measures helping to interpret the plots in PCA, FCA and in PTAk as well. The sum of the COS2 across all the components ()needed to rebuild fully the tensor analysed) would make 1000 and the sum pf the CTR across the entry mode would be 1000.

References

Escoufier Y (1985) L'Analyse des correspondances : ses propriétés{propretes} et ses extensions. ISI 45th session Amsterdam.

Leibovici D(1993) Facteurs à{a} Mesures Répétées{Repetees} et Analyses Factorielles : applications à{a} un suivi Epidémiologique{Epidemiologique}. Université{Universite} de Montpellier II. PhD Thesis in Mathématiques{Mathematiques} et Applications (Biostatistiques).

Leibovici DG (2010) Spatio-temporal Multiway Decomposition using Principal Tensor Analysis on k-modes:the R package PTAk. Journal of Statistical Software, 34(10), 1-34. http://www.jstatsoft.org/v34/i10/

See Also

PTAk, FCA2, FCAk, summary.FCAk, plot.PTAk

Examples

data(crimerate)
 cri.FCA2 <- FCA2(crimerate)
 summary(cri.FCA2)
  plot(cri.FCA2, mod = c(1,2), nb1 = 2, nb2 = 3) # unscaled
  plot(cri.FCA2, mod = c(1,2), nb1 = 2, nb2 = 3, coefi = list(rep(0.130787,50),rep(0.104359,7)) )# symmetric-map biplot
 CTR(cri.FCA2, mod = 1, solnbs = 2:4)
 CTR(cri.FCA2, mod = 2, solnbs = 2:4)
 COS2(cri.FCA2, mod = 2, solnbs = 2:4, FCA = TRUE)