Pumas version 0.11.0 (Alpha 11) released

Announcements
1

A new release of Pumas is now available in the JuliaProRegistry. You can upgrade your current installation of Pumas by simply running ] up Pumas at the command prompt in JuliaPro.

Release Notes for Pumas v0.11.0

Features

  • The LaplaceI approximation is now available when fitting models with multiple dependent variables
  • New coeftable function to produce a DataFrame of coefficients from a FittedPumasModel .
  • The vcov function no longer throws an exception when the computation fails. Instead, it produces a warning and returns a failed FittedPumasModelInference object.
  • Improvements to the compilation time for models with many random effects. The StaticArrays package is now used for vectors only whereas normal arrays are used for matrices.
  • It’s now possible to fit models with a @random block using the NaivePooled method.
  • Define an empirical_bayes_dist method to compute the empirical Bayes estimates with uncertainties for a fitted model.
  • Improve error message when all values of a covariate are missing for a subject.
  • Bootstrapping via a new bootstrap function.

Deprecations

  • The empirical_bayes function now computed the post hoc estimates when using the Pumas.FO approximation.
  • The NCAReport function now returns a DataFrame instead of an NCAReport struct.
  • The param property of FittedPumasModel has been deprecated in favor of coef(::FittedPumasModel)

Bugfixes

  • Fix DataFrame constructor for SimulatedObservations when the model doesn’t include events.
  • Change of steady-state event calculation from integer to floating point division to work around breaking changes for integer division in Julia 1.4.