README.md

In Wiedi97/LRCoDa-Package: Compositional Data Analysis with Linear Regression

{LRCoDa-Package}

Linear Regression to combine the use of compositional and non-compositional data in the same model (Planned as extension of the lmCoDaX function from the robComposition package)

using LRCoDa

data(gemas)

X <- dplyr::select(gemas, c(MeanTemp, soilclass, Al:Zr))

lr <- LRCoDa(y = gemas$sand, X, external = c('MeanTemp', 'soilclass'))

What is newly released regarding to the lmCoDaX function from the robComposition Package

• Use linear regression for compositional and non-compositional in the same model
• Include factor variable in the linear regression with compositional data
• Use lmrob for robust regression
• Can specify the norm to calculate the pivot coordinates
• Prefix Notation of the variables available in the output

Goals

• Extend the possibilities of using the linear regression for compositional data.

Getting Started

Dependencies

The package has dependencies on

R (>= 2.10), dplyr, robCompositions, robustbase, sjmisc

Installation

Installion of LRCoDa is easy when the R-tools are installed. Just use

library(devtools)
install_github("Wiedi97/LRCoDa-Package")

Wiedi97/LRCoDa-Package documentation built on June 30, 2022, 10:51 a.m.