Initial CRAN release
Bootstrap-calibrated GoF test for simplex regression
Paper reproduction functions included
First release.
simplex_fit(): maximum-likelihood fit of the simplex regression model with
constant or varying dispersion (logit/log links), with C++ kernels via Rcpp.
simplex_gof(): parametric bootstrap-calibrated goodness-of-fit test based on
the aggregated local-influence statistic U_n.
simplex_diag(): local-influence diagnostics, with a choice of analytic
(bootstrap-invariant) or finite-difference gradient.
simplex_Un_asymptotic(): asymptotic version of the statistic for the
null-distribution study.
Plotting: plot_influence(), plot_envelope() (bootstrap envelope),
plot_gof_boot(), plus plot() methods for the fitted and test objects.
Reproduction helpers: paper_ammonia(), paper_pbsc(), paper_fig1(), and
standalone scripts in inst/scripts/.
Two bundled datasets: ammonia (n = 21) and pbsc (n = 239).