modeLLtest 1.0.6
cvmf() now computes the cross-validated partial
log-likelihoods in-house rather than refitting
survival::coxph() with offset four times per obs. Results
are unchanged (differences might exist at the level of floating-point
rounding)- Registered the S3 print methods for cvdm, cvll, cvlldiff, and cvmf
objects in NAMESPACE. printing a result now shows the formatted output
instead of raw list.
- print.cvmf() now reports the robust estimator’s extended Wald test
(ewald.test from coxrobust::coxr()) on the “Extended Wald test” line; it
previously showed the partial-likelihood Wald statistic.
- Long calls now print on separate lines instead of being run
together.
- Added summary() methods for cvdm, cvll, cvlldiff, and cvmf objects,
each with matching print() method. They report the stored test results
with more detail (e.g., the test statistic, degrees of freedom, and
coefficient tables) and return a list for programmatic use.
modeLLtest 1.0.5
- Updated requirements for C++ (in makevars files) on request from
CRAN
- Added a note about the typo in the formula in the paper (but not the
code) into documentation for cvdm.R and cvll.R
- Made minor fixes, such as changing if() conditions when comparing
class() to string
modeLLtest 1.0.4
- Took out Travis and updated to gh-actions
- Made minor edits, including removing onload citation and cleaning
documentation
modeLLtest 1.0.3
- Updated typos in data files and cleaned documentation
modeLLtest 1.0.2
- Added a
NEWS.md file to track changes to the
package
- Updated documentation
modeLLtest 1.0.1
- JOSS-accepted release.
- Added minor changes to increase community access such as
CONTRIBUTING file, CODE_OF_CONDUCT file, and issues
modeLLtest 1.0.0