summary.PTAk {PTAk} | R Documentation |
Print a summary listing of the decomposition obtained.
summary.PTAk(object,testvar=1,dontshow="*",...) summary.FCAk(object,testvar=0.5,dontshow="*",...)
object |
an object inheriting from class PTAk , representing a generalised singular value decomposition |
testvar |
control within nTens used Principal Tensor with minimum percent of variability explained |
dontshow |
boolean criterion to remove Principal Tensors from the summary, or
default is a character "*" equivalent to the criterion:
!substr(solution[[length(solution)]][["vsnam"]],1,1)=="*" |
... |
summary generic additional arguments not used here |
The function prints a listing of the decomposition with historical
order (instead of traditional singular value order). It is useful
before any plots or reconstruction, a screeplot (using
plot.PTAk
) will be also useful. It is useful before any plots
r reconstruction, a screeplot (using plot.PTAk
) will be also
useful. summary.FCAk
is alike
summary.PTAk
but testvar
operates on the variability of
the lack of complete independence.
prints on the prompt
At the moment can be used for PCAn
,
CANDPRA
, better summaries will be in the next release.
Didier Leibovici c3s2i@free.fr
Leibovici D (2000) Multiway Multidimensional Analysis for Pharmaco-EEG Studies.(submitted) http://c3s2i.free.fr/cv/recentpub.html
data(crimerate) crimerate.mat <- sweep(crimerate,2,apply(crimerate,2,mean)) crimerate.mat <- sweep(crimerate.mat,2,sqrt(apply(crimerate,2,var)),FUN="/") cri.svd <- SVDgen(crimerate.mat) summary(cri.svd,testvar=0) plot(cri.svd,scree=TRUE) par(new=TRUE) RiskJackplot(cri.svd,nbvs=1:7,mod=NULL,max=NULL,rescaled=TRUE, axes=FALSE,ann=FALSE) par(new=FALSE) # or equivalently plot(cri.svd,scree=TRUE,type="b",lty=3,RiskJack=1) #set mod=NULL or c(1,2) ### data(crimerate) criafc <- FCAmet(crimerate,chi2=TRUE) cri.afc <- SVDgen(criafc$data,criafc$met[[2]],criafc$met[[1]]) summary(cri.afc) plot(cri.afc,scree=TRUE) plot(cri.afc,scree=TRUE,type="b",lty=3,RiskJack=1,method="FCA")