Name | Modified | Size | Downloads / Week |
---|---|---|---|
Parent folder | |||
README.md | 2025-08-26 | 3.2 kB | |
v5.0.0 source code.tar.gz | 2025-08-26 | 289.0 kB | |
v5.0.0 source code.zip | 2025-08-26 | 340.6 kB | |
Totals: 3 Items | 632.7 kB | 0 |
MixedModels v5.0.0
Version 5.0.0 contains some user-visible changes and many changes in the underlying code.
- Options related to multithreading in the bootstrap have been completely removed.
- Model fitting now uses unconstrained optimization, with a post-fit canonicalization step so that the diagonal elements of the lower Cholesky factor are non-negative. Relatedly, support for constrained optimization has been completely removed and the lowerbd
field of OptSummary
dropped.
- The default optimizer has changed to use NLopt's implementation of NEWUOA. Further changes to the default optimizer are considered non-breaking.
- The profile
function now respects backend and optimizer settings.
- The deprecated hide_progress
keyword argument has been removed in favor of the shorter and affirmative progress
.
- A fitlog is always kept and stored as a Tables.jl-compatible column table.
- See NEWS.md for more details
NB: Closed issues and pull requests are sorted temporally and so may
include backports to other versions or work in the development branch for
an upcoming breaking release. Please see the NEWS file
for changes sorted by release.
Closed issues:
- Consistency in progress options (#567)
- Use RegressionFormulae.jl for defining nesting syntax in fixed effects (#599)
- Simplify fitlog logic (#618)
- Always use -2 loglikelihood
instead of deviance
for LRT computations (#633)
- Drop multithreading in bootstrap (#675)
- Error in profile(m)
mutates m
and fails to clean up (#794)
- Should there be a method for predict(m::GenearlizedLinearMixedModel; type = ...)
(#817)
- export likelihoodratiotest
using an alias for non-nested tests (#818)
- Missing method in MixedModelsForwardDiffExt.jl (#845)
- Not able to run examples as can't find "dat.rda" file (#847)
- dataset
and datasets
has been removed (#852)
Merged pull requests:
- Add data and script for issue 833 (#835) (@dmbates)
- [WIP] Switch to unconstrained optimizers - step 1 (#840) (@dmbates)
- stopgap fix for Aqua.jl on Julia 1.13 nightly (#843) (@palday)
- Revert "stopgap fix for Aqua.jl on Julia 1.13 nightly (#843)" (#844) (@palday)
- Bump actions/checkout from 4 to 5 (#846) (@dependabot[bot])
- Deprecations, documentation and clean up in preparation for 5.0 (#848) (@palday)
- remove lowerbd field from OptSummary (#849) (@palday)
- replace fitlog with a column table, always populate it (#850) (@palday)
- Add timingtable.jl (#851) (@dmbates)
- optimization info/backend updates (#853) (@palday)
- remove dataset and datasets (#854) (@palday)
- remove local definitions of fulldummy and nesting operator (#855) (@palday)
- allow specifying type
argument in GLMM predict on original data (#856) (@palday)
- provide a breadcrumb when spline interpolation fails during profiling (#857) (@palday)
- rework likelihoodratio test to be just a pretty-printing wrapper around lrtest (#858) (@palday)