rrcov3way (version 0.1-18)

ulabor: Undeclared labor by region in Italy

Description

The dataset contains the undeclared labor in thousands work units. The data originate from Italy and are recorded at a regional level over a certain time horizon for five macroeconomic activities defined according to NACE Rev. 1.1 classification.

Usage

data("ulabor")

Format

A three-way array with dimension 22x5x5. The first dimension refers to 22 regions in Italy. The second dimension refers to the 5 economic activities. The third dimension refers to the years in the period 2001-2009.

References

ISTAT (2011). Note metodologiche, la misura dell'occupazione non regolare nelle stime di contabilita nazionale [online]. Roma. Available at: http://www.istat.it/it/archivio/39522.

Di Palma M.A., Filzmoser P., Gallo M. and Hron, K. (2016). A robust CP model for compositional data, submitted.

Examples

# NOT RUN {
data(ulabor)
dim(ulabor)
str(ulabor)

## Plot robust and non-robust DD-plots of the ilr-transformed data
usr <- par(mfrow=c(1,2))
res1 <- Parafac(ulabor, robust=TRUE, coda.transform="ilr")
res2 <- Parafac(ulabor, coda.transform="ilr")
plot(res1)
plot(res2)
par(usr)

# }
# NOT RUN {
## Plot Orthonormalized A-mode component plot
res <- Parafac(ulabor, robust=TRUE, coda.transform="ilr")
plot(res, which="comp", mode="A", main="Component plot, A-mode")

## Plot Orthonormalized B-mode component plot
plot(res, which="comp", mode="B", main="Component plot, B-mode")

## Plot Orthonormalized B-mode component plot
plot(res, which="comp", mode="C", main="Component plot, C-mode")


## Per component plot
## adapted for the example and only for robust, ilr transformed model
##
##
res <- Parafac(ulabor, robust=TRUE, coda.transform="ilr")

plot(res, which="percomp")              # component 1
plot(res, which="percomp", comp=2)      # component 2
# }