Non linear least squares regression for percentages or proportions
Non linear least squares regression for percentages or proportions.
propols.reg(y, x, cov = FALSE, tol = 1e-07 ,maxiters = 100)
y |
The dependent variable, a numerical vector with percentages or proporions, including 0s and or 1s. |
x |
A matrix with the indendent variables. |
cov |
Should the sandwich covariance matrix and the standard errors be returned? If yes, set this equal to TRUE. |
tol |
The tolerance value to terminate the Newton-Raphson algorithm. This is set to 10^{-7} by default. |
maxiters |
The maximum number of iterations that can take place during the fitting. |
The ordinary least squares between the observed and the fitted percentages is adopted as the objective function. This involves numerical optimization since the relationship is non-linear. There is no log-likelihood. This is the univariate version of the OLS regression for compositional data mentioned in Murteira and Ramalho (2016).
A list including:
sse |
The sum of squares of the raw residuals. |
be |
The beta coefficients. |
seb |
The sandwich standard errors of the beta coefficients, if the input argument argument was set to TRUE. |
covb |
The sandwich covariance matrix of the beta coefficients, if the input argument argument was set to TRUE. |
iters |
The number of iterations required by the Newton-Raphson algorithm. |
Michail Tsagris
R implementation and documentation: Michail Tsagris mtsagris@uoc.gr.
Murteira, Jose MR, and Joaquim JS Ramalho 2016. Regression analysis of multivariate fractional data. Econometric Reviews 35(4): 515-552.
y <- rbeta(150, 3, 4) x <- iris a <- propols.reg(y, x)
Please choose more modern alternatives, such as Google Chrome or Mozilla Firefox.