Robust Methods for Multiway Data Analysis, Applicable also for Compositional Data.
rrcov3way: Robust Methods for Multiway Data Analysis
The package provides robust methods for multiway data analysis by means of Parafac and Tucker 3 models (Engelen and Hubert (2011) doi:10.1016/j.aca.2011.04.043). Robust versions for compositional data are also provided (Gallo (2015) doi:10.1080/03610926.2013.798664, Di Palma et al. (2018) doi:10.1080/02664763.2017.1381669).
Several optimization methods alternative to ALS are available (Simonacci and Gallo (2019) doi:10.1016/j.chemolab.2019.103822, Simonacci and Gallo (2020) doi:10.1007/s00500-019-04320-9).
Installation
The rrcov3way package is on CRAN (The Comprehensive R Archive Network) and the latest release can be easily installed using the command
install.packages("rrcov3way")
Building from source
To install the latest stable development version from GitHub, you can pull this repository and install it using
## install.packages("remotes")
remotes::install_github("valentint/rrcov3way")
Of course, if you have already installed remotes, you can skip the first line (I have commented it out).
Example
This is a simple example which shows you basic functions of the package using the OECD elind data set. The data consist of specialization indices of electronics industries of 23 European countries for the years 1973-1979. The specialization index is defined as the proportion of the monetary value of an electronic industry compared to the total export value of manufactured goods of a country compared to the similar proportion for the world as a whole.
## Load the package 'rrcov3way' and the data set
## to be used in the examples:
##
## OECD Electronics Industries Data - Kroonenberg PM (2008)
## 23 countries x 6 industries x 7 years
library(rrcov3way)
#> Robust Methods for Multiway Data Analysis, Applicable also for
#> Compositional Data (version 0.2-5)
#>
#> Attaching package: 'rrcov3way'
#> The following object is masked from 'package:stats':
#>
#> reorder
data(elind)
dim(elind)
#> [1] 23 6 7
## The frontal slices (mode C) are matrices with dimension 23x6,
## representing the data for all countries and all industries
## in one year. The labels are:
rownames(elind[,,1])
#> [1] "CA" "US" "JP" "AS" "NZ" "BL" "DA" "FR" "RF" "GR" "IR" "IT" "PB" "RU" "AU"
#> [16] "FI" "NO" "PO" "SP" "SV" "CH" "TU" "YU"
colnames(elind[,,1])
#> [1] "INFO" "RADI" "TELE" "STRU" "ELET" "COMP"
## The lateral slices (mode B) represent the data
## for all countries and years for one industry
## - matrices with dimension 23 x 7 and the labels are:
rownames(elind[,1,])
#> [1] "CA" "US" "JP" "AS" "NZ" "BL" "DA" "FR" "RF" "GR" "IR" "IT" "PB" "RU" "AU"
#> [16] "FI" "NO" "PO" "SP" "SV" "CH" "TU" "YU"
colnames(elind[,1,])
#> [1] "78" "79" "80" "82" "83" "84" "85"
## First of all we center and scale the data, using the default procedures
## for centering and scaling.
elind <- do3Scale(elind, center=TRUE, scale=TRUE)
## Next we perform classical PARAFAC analysis
## with the default number of components (ncomp=2).
## default PARAFAC, non-robust, no ilr transformation,
## extract 3 components
res <- Parafac(elind, ncomp=3)
res
#> Call:
#> Parafac(X = elind, ncomp = 3)
#>
#>
#> PARAFAC analysis with 3 components.
#> Fit value: 2.522052
#> Fit percentage: 57.97 %
#>
Community guidelines
Report issues and request features
If you experience any bugs or issues or if you have any suggestions for additional features, please submit an issue via the Issues tab of this repository. Please have a look at existing issues first to see if your problem or feature request has already been discussed.
Contribute to the package
If you want to contribute to the package, you can fork this repository and create a pull request after implementing the desired functionality.
Ask for help
If you need help using the package, or if you are interested in collaborations related to this project, please get in touch with the package maintainer.