optimTP.GA() now directly uses function kofnGA::kofnGA() and supports parallel computation of the
objective function via the cluster argument.lrtest() / lrtest.twoPhaseSPML() — likelihood ratio test for two nested
twoPhaseSPML models, mirroring the interface of lmtest::lrtest().
The function computes LRT = -2 * (ll0 - ll1) and compares it to a
chi-squared distribution whose degrees of freedom equal the difference in
the number of estimated regression coefficients. Model order (null vs
alternative) is determined automatically. Returns an object of class
"lrtest.twoPhaseSPML" with a dedicated print method that displays a
formatted table analogous to lmtest::lrtest() output.
twoPhaseSPML() now only takes a single data argument in line with most
implementations.
DataGeneration_TPD() now generates data from any member of the supported
glm families.
twoPhaseSPML() now returns an object of S3 class "twoPhaseSPML" rather
than a plain list. The returned object additionally contains the fitted
formula, miscov, family, and a null.model flag.
print.twoPhaseSPML() — a glm-style print method for objects returned by
twoPhaseSPML(). Displays the call, a coefficient table (with standard
errors and p-values under the alternative model), family/link, dispersion
(Gaussian/Gamma only), log-likelihood, EM iteration count, and the relevant
hypothesis-test statistic with its p-value.
T / F shorthand with TRUE / FALSE throughout all source
files.class(x) == "..." comparisons with is.*() / inherits()
calls.print() used for a diagnostic message in twoPhaseSPML() with
message().utils from Depends to Imports and tightened the Depends field
to list only R itself.DESCRIPTION: used Authors@R, added URL and BugReports
fields, expanded the package description.NAMESPACE to register print.twoPhaseSPML as an S3 method via
S3method(print, twoPhaseSPML).