Changelog
Source:NEWS.md
missingHE 1.6.3
Updated documentation across all function files, datasets, and
README.md.Updated package metadata,
inst/CITATION, and academic references (DESCRIPTION,.Rdfiles, andinst/REFERENCES.bib).Standardised canonical CRAN URLs and DOI links across documentation.
Updated all package functions to remove redundant code lines and add a selection of test functions.
Added new unit tests (
testthat) covering all key package functions.Updated the version of the package in the
DESCRIPTIONfile to1.6.3.
missingHE 1.6.2
CRAN release: 2026-10-13
Creation of the new S3 method function
postsumto extract and display posterior summaries about model parameters filtered based on some optional arguments.Update of the
print.missingHEfunction to return more standardised output about model coefficient estimates following usual practice forprintfunctions inR.Code for some functions slightly updated to remove specification of deprecated arguments no longer valid in the most recent versions of some dependence packages’ functions.
Update of the output generated using the
printfunction.The package vignettes have been updated.
Updated the version of the package in the
DESCRIPTIONfile to1.6.2.
missingHE 1.6.1
CRAN release: 2026-06-01
1 June 2026
Bug fixes that prevented the correct implementation of the models depending on specific combination of distributional assumptions.
Code for some functions slightly updated to ensure coherence among prior values assumed across different model specifications.
Update of some default prior values to cover a less broad range of values.
The package vignettes have been updated.
Updated the version of the package in the
DESCRIPTIONfile to1.6.1.
missingHE 1.6
19 March 2026
A few package functions have been updated to allow specification of models for multiple treatment arm trials and produce output which allows a higher degree of customisation, such as generating graphical and numeric results by treatment, time point and for different types of estimates of interest.
Key functions that were updated include:
selection,pattern,hurdleandlmdm(previously calledlong_miss).Code for many functions dedicated to the fitting of models and extracting model results has been updated and more efficient.
Key functions that were updated include:
ppc,pic,plot.missingHE,coef.missingHEandsummary.missingHENew customisation options have been incorporated into the package functions to allow a wider choice in terms of output generation and display.
The package vignettes have been updated and extended to provide a more general and detailed presentation of the main package functions using the built-in data sets as motivating cases.
The package built-in data sets (
MenSS.RDataandPBS.RData) have been updated to align variable names and types to match the requirements of the inputs accepted by the updated functions.Function used to fit longitudinal model has been removed and replaced with a new one (
lmdm) to provide higher degree of flexibility and more customisation options for the user when fitting the model, including assumptions about time dependence, missingness patterns and specification of priors.Available choices in terms of prior distribution customisation and modification have been increased for all models through a more intuitive approach to prior specification with examples provided in the package vignettes.
New types of posterior predictive checks (
ppc), information criteria (pic), imputed data graphs (plot.missingHE) and diagnostics (diagnostic) have been added to the pool of available choices in the corresponding functions.The generation of the quantities/graphs has been updated through new function arguments and customised based on the type of model fitted.
Updated the version of the package in the
DESCRIPTIONfile to1.6.0.