[R-sig-ME] lme4 1.1-19 going to CRAN soon

[Ben Bolker]

This is a heads-up that we'll be sending a new version of lme4 to CRAN
soon. This is not a major revision (the proximal reason for sending a
new version is to address some minor CRAN-compatibility issues).
However, I did *almost* make a "brown paper bag release"
<http://www.catb.org/jargon/html/B/brown-paper-bag-bug.html>, so if
anyone feels like taking the development version for a spin in the next
day or so that would be appreciated. (The new version is due on CRAN by
Nov 9; we have otherwise completed testing on Travis, win-builder,
reverse dependencies, etc ...)

cheers
Ben Bolker


=====
NEW FEATURES:

• influence measure code from ‘car’ rolled in (see
‘?influence.merMod’)

• ‘mkReTrm’ gets new arguments ‘reorder.terms’, ‘reorder.vars’
to control arrangement of RE terms and individual effects
with RE terms within model structures

• adding material from the ‘RePsychLing’ package (on GitHub;
see Bates et al 2015 arXiv:1506.04967) to show orthogonal
variance components.

USER-VISIBLE CHANGES:

• ‘as.data.frame.merMod’ finds conditional variance
information stored either as ‘attr(.,"postVar")’ or
‘attr(.,"condVar")’ (for ‘glmmTMB’ compatibility)

• change to defaults of ‘[g]lmerControl’ to print a message
when fits are singular

• new utility ‘isSingular()’ function for detecting singular
fits; by default a message is printed for singular fits

• ‘allFit’ function/methods have been moved to the main
package, rather than being included in an auxiliary source
file

• post-fitting convergence checks based on estimated gradient
and Hessian (see ‘troubleshooting’) are no longer performed
for (nearly-)singular fits (see ‘isSingular’)