MAINT.Data (version 1.0.1)

da-methods: Discriminant Analysis of Interval Data

Description

lda and qda perform linear and quadratic discriminant analysis of Interval Data based on classic estimates of a mixture of Gaussian models. Roblda and Robqda do the same using robust estimates of location and scatter. snda performs discriminant analysis of Interval Data based on estimates of mixtures of Skew-Normal models

Usage

# S4 method for IData
lda( x, grouping, prior="proportions", CVtol=1.0e-5, egvtol=1.0e-10, 
  subset=1:nrow(x), CovCase=1:4, SelCrit=c("BIC","AIC"), silent=FALSE, … )
# S4 method for IdtMxtNDE
lda(x, prior="proportions", selmodel=BestModel(x), egvtol=1.0e-10,
  silent=FALSE, … )
# S4 method for IdtClMANOVA
lda( x, prior="proportions", selmodel=BestModel(H1res(x)),
  egvtol=1.0e-10, silent=FALSE, … )
# S4 method for IdtClMANOVA
lda( x, prior="proportions", selmodel=BestModel(H1res(x)),
  egvtol=1.0e-10, silent=FALSE, … )
# S4 method for IdtLocNSNMANOVA
lda( x, prior="proportions", 
  selmodel=BestModel(H1res(x)@NMod), egvtol=1.0e-10, silent=FALSE, … )
# S4 method for IData
qda( x, grouping, prior="proportions", CVtol=1.0e-5, subset=1:nrow(x),
  CovCase=1:4, SelCrit=c("BIC","AIC"), silent=FALSE, … )
# S4 method for IdtMxtNDE
qda(x, prior="proportions", selmodel=BestModel(x), silent=FALSE, 
  … )
# S4 method for IdtHetNMANOVA
qda( x, prior="proportions", selmodel=BestModel(H1res(x)), 
  silent=FALSE, … )
# S4 method for IdtGenNSNMANOVA
qda( x, prior="proportions", 
  selmodel=BestModel(H1res(x)@NMod), silent=FALSE, … )
# S4 method for IData
Roblda( x, grouping, prior="proportions", CVtol=1.0e-5, egvtol=1.0e-10,
  subset=1:nrow(x), CovCase=1:4, SelCrit=c("BIC","AIC"), silent=FALSE, 
  CovEstMet=c("Pooled","Globdev"), SngDMet=c("fasttle","fulltle"),
  Robcontrol=RobEstControl(), … )
# S4 method for IData
Robqda( x, grouping, prior="proportions", CVtol=1.0e-5, 
  subset=1:nrow(x), CovCase=1:4, SelCrit=c("BIC","AIC"), silent=FALSE,
  SngDMet=c("fasttle","fulltle"), Robcontrol=RobEstControl(), … )
# S4 method for IData
snda(x, grouping, prior="proportions", CVtol=1.0e-5, subset=1:nrow(x),
  CovCase=1:4, SelCrit=c("BIC","AIC"), Mxt=c("Loc","Gen"), … )
# S4 method for IdtLocSNMANOVA
snda( x, prior="proportions", selmodel=BestModel(H1res(x)),
  egvtol=1.0e-10, silent=FALSE, … )
# S4 method for IdtLocNSNMANOVA
snda( x, prior="proportions",
  selmodel=BestModel(H1res(x)@SNMod), egvtol=1.0e-10, silent=FALSE, … )
# S4 method for IdtGenSNMANOVA
snda( x, prior="proportions", selmodel=BestModel(H1res(x)),
  silent=FALSE, … )
# S4 method for IdtGenNSNMANOVA
snda( x, prior="proportions",
  selmodel=BestModel(H1res(x)@SNMod), silent=FALSE, … )

Arguments

x

An object of class , , , or with either the original Interval Data, or the results of a Interval Data Skew-Normal MANOVA, from which the discriminant analysis will be based.

grouping

Factor specifying the class for each observation.

prior

The prior probabilities of class membership. If unspecified, the class proportions for the training set are used. If present, the probabilities should be specified in the order of the factor levels.

CVtol

Tolerance level for absolute value of the coefficient of variation of non-constant variables. When a MidPoint or LogRange has an absolute value within-groups coefficient of variation below CVtol, it is considered to be a constant.

subset

An index vector specifying the cases to be used in the analysis.

CovCase

Configuration of the variance-covariance matrix: a set of integers between 1 and 4.

SelCrit

The model selection criterion.

silent

A boolean flag indicating wether a warning message should be printed if the method fails.

CovEstMet

Method used to estimate the common covariance matrix in Roblda (Robust linear discriminant analysis). Alternatives are “Pooled” (default) for a pooled average of the the robust within-groups covariance estimates, and “Globdev” for a global estimate based on all deviations from the groups multivariate l1 medians. See Todorov and Filzmoser (2009) and pcaPP.l1median for details.

SngDMet

Algorithm used to find the robust estimates of location and scatter. Alternatives are “fasttle” (default) and “fulltle”.

Robcontrol

A control object (S4) of class RobEstControl-class containing estimation options - same as these provided in the function specification. If the control object is supplied, the parameters from it will be used. If parameters are passed also in the invocation statement, they will override the corresponding elements of the control object.

Mxt

Indicates the type of mixing distributions to be considered. Current alternatives are “Hom” (homocedastic) and “Het” (hetereocedasic) for Gaussian models, “Loc” (location model -- groups differ only on their location parameters) and “Gen” “Loc” (general model -- groups differ on all parameters) for Skew-Normal models.

selmodel

Selected model from a list of candidate models saved in object x.

egvtol

Tolerance level for the eigenvalues of the product of the inverse within by the between covariance matrices. When a eigenvalue has an absolute value below egvtol, it is considered to be zero.

…

Other named arguments.

References

Duarte Silva, A.P. and Brito, P. (2015), Discriminant analysis of interval data: An assessment of parametric and distance-based approaches. Journal of Classification 39(3), 516--541.

Todorov V. and Filzmoser P. (2009), An Object Oriented Framework for Robust Multivariate Analysis. Journal of Statistical Software 32(3), 1--47.

See Also

, ,,,, pcaPP.l1median.

Examples

# Create an Interval-Data object containing the intervals for 899 observations 
# on the temperatures by quarter in 60 Chinese meteorological stations.

ChinaT <- IData(ChinaTemp[1:8],VarNames=c("T1","T2","T3","T4"))

#Linear Discriminant Analysis

ChinaT.lda <- lda(ChinaT,ChinaTemp$GeoReg)
cat("Temperatures of China -- linear discriminant analysis results:\n")
print(ChinaT.lda)
cat("lda Prediction results:\n")
print(predict(ChinaT.lda,ChinaT)$class)


#Estimate error rates by ten-fold cross-validation replicated 20 times  

CVlda <- DACrossVal(ChinaT,ChinaTemp$GeoReg,TrainAlg=lda,CovCase=CovCase(ChinaT.lda))
summary(CVlda[,,"Clerr"])
glberrors <- 
  apply(CVlda[,,"Nk"]*CVlda[,,"Clerr"],1,sum)/apply(CVlda[,,"Nk"],1,sum)
cat("Average global classification error =",mean(glberrors),"\n")


#Quadratic Discriminant Analysis

ChinaT.qda <- qda(ChinaT,ChinaTemp$GeoReg)
cat("Temperatures of China -- qda discriminant analysis results:\n")
print(ChinaT.qda)


#Estimate error rates by ten-fold cross-validation replicated 20 times  

CVqda <- DACrossVal(ChinaT,ChinaTemp$GeoReg,TrainAlg=qda,CovCase=CovCase(ChinaT.qda))
summary(CVqda[,,"Clerr"])
glberrors <- 
  apply(CVqda[,,"Nk"]*CVqda[,,"Clerr"],1,sum)/apply(CVqda[,,"Nk"],1,sum)
cat("Average global classification error =",mean(glberrors),"\n")

# Skew-Normal based discriminant analysis, asssuming that the different regions may differ
# in all SkewNormal parameters

cat("Temperatures of China -- SkewNormal general model discriminant analysis results:\n")
ChinaT.gensnda <- snda(ChinaT,ChinaTemp$GeoReg,Mxt="Gen")
print(ChinaT.gensnda)

#Estimate error rates by three-fold cross-validation without replication  

CVgensnda <- DACrossVal(ChinaT,ChinaTemp$GeoReg,TrainAlg=snda,Mxt="Gen",
  CovCase=CovCase(ChinaT.gensnda),kfold=3,CVrep=1)
summary(CVgensnda[,,"Clerr"])
glberrors <- 
  apply(CVgensnda[,,"Nk"]*CVgensnda[,,"Clerr"],1,sum)/apply(CVgensnda[,,"Nk"],1,sum)
cat("Average global classification error =",mean(glberrors),"\n")

#Robust Quadratic Discriminant Analysis

ChinaT.rqda <- Robqda(ChinaT,ChinaTemp$GeoReg)
cat("Temperatures of China -- robust qda discriminant analysis results:\n")
print(ChinaT.rqda)

#Estimate error rates by ten-fold cross-validation with 5 replications 

CVrqda <- DACrossVal(ChinaT,ChinaTemp$GeoReg,TrainAlg=Robqda,CovCase=CovCase(ChinaT.rqda),
   CVrep=5)
summary(CVrqda[,,"Clerr"])
glberrors <- 
  apply(CVrqda[,,"Nk"]*CVrqda[,,"Clerr"],1,sum)/apply(CVrqda[,,"Nk"],1,sum)
cat("Average global classification error =",mean(glberrors),"\n")