Fixed a platform-specific rounding difference in the ‘lrstat’ sample
size test that caused a failure on macOS (ARM64). The test now accepts
the correct result on both Windows and macOS.
survSampleSize 0.1.1
Addressed CRAN review comments:
Software and package names (‘shiny’, ‘lrstat’, ‘powerSurvEpi’) are
now written in single quotes in the Description.
Added ‘testthat’ unit tests for the internal (non-exported)
calculation functions, so that breaking changes in R or in the upstream
packages are detected automatically.
Moved ‘lrstat’ and ‘powerSurvEpi’ from Suggests to Imports, as they
are now used directly by the package’s internal calculation engine.
survSampleSize 0.1.0
Initial release.
run_app() launches an interactive Shiny application for
survival-endpoint sample size and power calculation.
Implements the Lu (2021) weighted log-rank method (via
lrstat) and the Freedman (1982) method (via
powerSurvEpi), with survival-curve and event-prediction
visualisations and a method comparison.