Constrained linear least squares for compositional responses and predictors.
ols.compcomp(y, x, rs = 5, tol = 1e-4, xnew = NULL)A matrix with the compositional data (dependent variable). Zero values are allowed.
A matrix with the compositional predictors. Zero values are allowed.
The number of times to run the constrained optimisation using different random starting values each time.
The threshold upon which to stop the iterations of the constrained optimisation.
If you have new data use it, otherwise leave it NULL.
A list including:
The time required by the regression.
The mean squared errors.
The beta coefficients.
The fitted of xnew if xnew is not NULL.
The function performs least squares regression where the beta coefficients are constained to be positive and sum to 1. We were inspired by the transformation-free linear regression for compositional responses and predictors of Fiksel, Zeger and Datta (2020).
Jacob Fiksel, Scott Zeger and Abhirup Datta (2020). A transformation-free linear regression for compositional outcomes and predictors. https://arxiv.org/pdf/2004.07881.pdf
# NOT RUN {
library(MASS)
set.seed(1234)
y <- rdiri(214, runif(4, 1, 3))
x <- as.matrix(fgl[, 2:9])
x <- x / rowSums(x)
mod <- ols.compcomp(y, x, rs = 1)
mod
# }
Run the code above in your browser using DataLab