pams 0.2.0
User interface
BootSmacof() now returns an object of class
pams_fit while retaining all previously available
named-list components.- Added
plot.pams_fit(), summary.pams_fit(),
print.pams_fit(), and print.summary.pams_fit()
methods.
- Corrected the documented and implemented default transformation to
type = "ordinal".
Reliability and
documentation
- Added systematic validation for data, profile count, directions,
confidence level, bootstrap count, participant indices, labels, and
choice arguments.
- Added an explicit warning when fewer than 1,000 bootstrap samples
are used.
- Made BCa calculations robust to boundary bias estimates and zero
jackknife acceleration denominators.
- Applied sign alignment to both bootstrap and jackknife
configurations.
- Clarified that person weights are unstandardized no-intercept OLS
coefficients and that
corDim values are partial
correlations.
- Added explicit documentation of sign indeterminacy and the absence
of general rotational or dimension-permutation alignment.
- Added a worked vignette, automated tests, and a multi-platform
GitHub Actions R CMD check workflow.
- Updated the maintainer address to
sekangandroid@gmail.com.
- Added optional vignette and tidy-tabular dependencies under
Suggests.