Changes in version 1.0.26 NEW FEATURES o confint method gains a std.err argument, for use when Wald standard errors are unavailable or unreliable o mle2 gains a hessian.method argument, which can be set to "optimHess" when the default method (hessian or jacobian from the numDeriv package) is unreliable Changes in version 1.0.25.1 (2023-12-09) CRAN COMPATIBILITY o remove/ignore spurious .Rout files Changes in version 1.0.25 (2022-05-11) USER-VISIBLE CHANGES o mle2.options() function is now exported CRAN COMPATIBILITY o vignettes will build without suggested packages Changes in version 1.0.24 (2021-08-05) USER-VISIBLE CHANGES o new namedrop_args argument to mle2 (don't drop argument names within sub-lists of data) o allow user-specified covariance matrix for importance sampling Changes in version 1.0.23 o suppress some warnings in tests for cross-platform/CRAN compatibility Changes in version 1.0.22 (2019-12-19) BUG FIXES o intercept-only parameters were handled wrong (GH #8) o fix error-handling (class vs. inherits) for R 4.0 compatibility NEW FEATURES o pop_pred_samp to compute population prediction samples, tweaking non-positive-definite covariance matrices and using importance samples if necessary/specified. *This is still VERY experimental/in flux!* o added dnorm_n function for Normal densities with the standard deviation parameter profiled out o add slnorm function for log-Normal densities (median and mean only for now) o IC tables now have an as.data.frame method (from @iago-pssjd on GitHub) USER-VISIBLE CHANGES o the evaluated start and fixed components of the call are now stored in the call component (Daniel B. Stouffer) Changes in version 1.0.20 (2017-10-30) BUG FIXES o fixed buglet: flipped profile plot axes, confint for negative values USER-VISIBLE CHANGES o summary.mle2 is now exported, for use by other packages o small fixes to AIC(c) methods Changes in version 1.0.19 (2017-04-18) o fixed bug: evaluate call$method so that profiling/updating works within a function environment o make AICtab smarter about NA values o fix BIC bug (infinite recursion) o hessian computation uses gradient function if provided o basic continuation method implemented for profiling (with stubs for smarter methods) o mle2 stores its calling environment for more flexibility when re-evaluating, e.g. in profiling (could lead to occasional surprises, e.g. if saving a fitted mle2 object with large objects in its calling environment) Changes in version 1.0.18 (2016-02-11) o update slice functionality; allow for explicit ranges o CRAN updates (import from base packages) Changes in version 1.0.17 (2014-05-27) o new warning if ~dnorm is called with sd implicitly ==1 o some internal tweaking to slice functionality o updated to allow for MuMIn v 1.10.0 changes Changes in version 1.0.16 (2014-03-02) o fix mnames behaviour o slight vignette cleanup Changes in version 1.0.15 (2013-11-21) o add logLik option to IC tables, more use of nobs methods o minor improvements to slice functionality Changes in version 1.0.14 o more CRAN tweaks Changes in version 1.0.13 (2013-08-23) o .Rbuildignore tweaks for CRAN Changes in version 1.0.12 o vignette tweaks o add Depends: R>=3.0.0 Changes in version 1.0.11 (2013-08-20) o add .Rbuildignore for CRAN tests Changes in version 1.0.10 o adapt to updated optimx o tweaks for CRAN compliance o adapt to updated optimx o tweaks for CRAN compliance Changes in version 1.0.9 o switch from aod to aods3 in vignettes o switch vignette to knitr o switch from aod to aods3 in vignettes o switch vignette to knitr Changes in version 1.0.8 o tweaks to print.ICtab() o tweaks to print.ICtab() Changes in version 1.0.7 o warn on convergence failure o warn on convergence failure Changes in version 1.0.6 o fixed (fairly obscure) bug in assigning data environment to gradient function; replace 'gr' in call with appropriately evaluated version o fixed (fairly obscure) bug in assigning data environment to gradient function; replace 'gr' in call with appropriately evaluated version Changes in version 1.0.5.3 o changed some cat() to message() o changed some cat() to message() Changes in version 1.0.5.2 (2012-08-01) o add .Rbuildignore for CRAN tests o remove BIC definitions (now unnecessary/should be adequately defined in core R) o add explicit Depends: on stats4 o note that development optimx (2012.05.24+) now fails on profiling when reduced model is 1-dimensional (i.e. for a 2-parameter model) Changes in version 1.0.5.1 (2012-07-17) o remove spurious .RData file; version bump Changes in version 1.0.5 o wrapped eigen() call to prevent failure of eigenvalue ratio summary when hessian is bad o fix bug: forgot to export summary method for mle2 o add exclude.fixed argument to coef method o fix bug: single lower/upper parameter, or prof.lower/prof.upper parameter, not interpreted properly in profile/confint o add and document slice methods: change from old (profile-like) behavior, old version is available as sliceOld o DESCRIPTION/NAMESPACE fixes, move most Depends: to Imports: instead (except methods package) Changes in version 1.0.4.2 o fix bug in gradient/vecpar/profiling interaction (Daniel Kaschek) o improve (and document) uniroot method for confint: now respects box constraints o fix issue where bobyqa (optimx) strips parameter names from result Changes in version 1.0.4.1 (2012-01-28) o remove a bit of installed junk; vignette fix Changes in version 1.0.4 (2012-01-02) o fix imports for latest version of R-devel Changes in version 1.0.3 (2011-09-14) o modified starting value code and documentation slightly: now allows per-parameter lists. Updated docs. o Fixed bug that would screw things up if 'i' were used as a parameter. o Update vignette to conform to MuMIn changes Changes in version 1.0.2 (2011-09-07) o fixed buglet in detection of no-intercept models when setting starting values Changes in version 1.0.1 o Turned off some warnings when skip.hessian=TRUE (e.g. in profile) o Calculate max grad and Hessian eigenval ratio, add to "details" o (Should add accessor methods) o Fixed NAMESPACE to export methods etc etc properly; removed universal exportPattern directive Changes in version 1.0.0 (2011-06-27) o mainly just a version bump for CRAN o added tag for variables in profile/confint warnings o lots more technical information in vignette about profiling algorithm Changes in version 0.9.9 o changed NEWS to NEWS.Rd o fixed bug for "optimize" - profile etc. now respect bounds o eliminated warning about bounds names if all identical o add "try_harder" flag to profiling (ignore flat spots, NAs ...) Changes in version 0.9.8 o gradient functions work better with fixed parameters, hence with profiling o profile plot reverts to linear for non-monotonic profile o added warning in confint for non-monotonic profile, or for non-monotonic spline fit to monotonic profile; revert from spline+linear to linear approximation in this case o various documentation improvements o optimx improvements o require data= argument when using formula interface o turn off hessian computation in profile o allow use of MASS::ginv Changes in version 0.9.7 (2011-01-10) o bug fix in calc_mle2_function for no-intercept models (thanks to Colin Kremer) o fixed optimx, added 'user' option Changes in version 0.9.6 o changed hessian calculation to use numDeriv code (causes tiny changes to hessian results that could matter in edge cases). Too lazy to provide a backward compatibility mode ... o documented optimizer= choices in ?mle2 Changes in version 0.9.5.1 (2010-05-22) o fixed bug in AICc (David Harris) Changes in version 0.9.5 o added NAMESPACE, various fixes to go with that o beginnings of an RUnit testing framework o tweaked vignette o added prof.lower, prof.upper to profile() o added "optimize" to list of allowed optimizers, some bug fixes Changes in version 0.9.4.1 o tweaked par() resetting in profile plots Changes in version 0.9.4 o more qAICc fixing Changes in version 0.9.3 (2009-09-18) o tweaked handling of bounds: profile now succeeds on some 1D problems where it didn't before o added deviance, residuals methods o added newparams argument to predict, simulate; newdata argument to simulate o added vignette (stub) o added explicit params argument, to help sort out full parameter specifications when parameters is non-NULL Changes in version 0.9.2 o fixed predict() for case with parameters o added snorm o changed ICtab defaults to weight=TRUE, base=FALSE, sort=TRUE Changes in version 0.9.1 o added simulate method (formula interface only) o fix AICctab bug o remove spurious cat/print in profile o fix qAIC bug Changes in version 0.9.0 (2009-02-04) o fix Tom Hobbs bug: named lower/upper/parscale/ndeps get rearranged properly, otherwise rearrange in order of "start" and issue a warning o documentation tweak for S4 as.data.frame o added sbeta to list of known distributions o removed nlme requirement & auto-loading Changes in version 0.8.9 (2008-08-07) o version bump, submit to CRAN o added predict method Changes in version 0.8.8 o added flexibility for profile plotting (main, x labels etc.); added examples o added an instance of "namedrop" to fix naming problem o added tol.newmin to slice etc. o added check for numeric return from profile within confint o fixed bugs in profile plotting when profile is restricted to a subset of variables o added tests for par() to reset to original on exit o improved profile documentation o replicate std.err if specified in profile o add as.data.frame o tweak tol.newmin (better fit found during profile) code Changes in version 0.8.7 o version bump, moved to R-forge. o reordered NEWS file (most recent first) Changes in version 0.8.6.1 o tweaked stop-on-better-fit code o fixed (?) qAIC(c) methods Changes in version 0.8.6 o tweak/fix to ICtab documentation (thanks to Tom Hobbs) o added qAIC(c) methods (not working yet!) Changes in version 0.8.5.1 o oops. Fixed infelicity (bug?) in new environment manipulation Changes in version 0.8.5 (2008-03-07) o tweaked environment/data assignment to preserve original minuslogl environment better Changes in version 0.8.4 o changed plot.profile.mle2 options (added onepage etc., made plot.confstr=TRUE by default) Changes in version 0.8.3 o added warning about too-short lower/upper o added documentation Changes in version 0.8.2 (2007-11-16) o fixed bug in AICctab o cosmetic change to printing - save call.orig o moved ChangeLog to NEWS Changes in version 0.8.1 (2007-10-22) o tried to use built-in relist, but failed: renamed relist to "relist2" (try again later) o documented get.mnames (auxiliary function for ICtabs) o started to add gr (gradient) capability - NOT TESTED Changes in version 0.8 o changed ICtab to allow either ICtab(x,y,z) or ICtab(list(x,y,z)) (L <- list(...); if is.list(L[[1]]) && length(L)==1) Changes in version 0.7.7 (2007-07-29) o fix bug in profiling: all optim() methods EXCEPT L-BFGS-B. return the value of the objective function if given a function with no arguments/zero-length starting parameter vector (this is the situation with "profiling" a 1-D function). L-BFGS-B gives funky answers. added a check for this case. (may need to check behavior for alternate optimizers (nlm etc)) [this behavior triggered a "found better fit" error when profiling 1D functions with L-BFGS-B] o changed behavior when finding better fit during profiling to return new parameters Changes in version 0.7.6 o tweak vignette o fixed second major AICc bug (was fixed in mle2 method, but not in logLik method) Changes in version 0.7.5 (2007-06-17) o change "ll" to "LL" in examples for clarity o tweaked anova reporting of models (wrap instead of truncating) o added (undocumented) show.points option to profile plot to display actual locations of profile evaluation o tweaked profile to behave better when profiling variables with constraints (upper, lower) o moved vignette to inst/doc where it belongs o ICtab hack to protect against package:aod definition of AIC(logLik) o added submit stub o tweaked slice.mle2-class docs for consistency o fiddled with vignette o preliminary code to allow non-monotonic profiles o preliminary add nlm to list of optimizers (untested) o add aod, Hmisc, emdbook to VignetteDepends and Suggests: Changes in version 0.7 (2007-05-21) o better df extraction in ICtab o minor bug fix for AICc (allows AICc of nls objects) o handle models with -1 in formula better: starting values set "all equal" o made ANOVA formula line-length accessible o added skip.hessian and trace arguments to mle2 o messed around with BIC definition - attempt at consistency with nlme o added rudimentary support for nlminb, constrOptim o nlme now required for fdHess (which is required for nlminb since it doesn't compute a finite-diff Hessian) Changes in version 0.6 (2007-02-18) o add experimental formula interface o change all names from mle to mle2 to avoid confusion/conflicts o with stats4 version of mle o change internal structure of data evaluation o worked on vignette o added optimizer slot (stub) Changes in version 0.5 o fix AICc bug! (was deviance+2*k*(k+1)/(n-k-1), not AIC+2*k*(k+1)/(n-k-1)) Changes in version 0.4 o change AIC to AICc for corrections o add AICtab for weights, delta, sort ... options o expose error messages occuring within profile() o uniroot tries harder to find a valid endpoint o truncate terms in anova.mle at 80 characters Changes in version 0.3 o enhanced anova method, works with print.anova o tweaked namedrop() code - ?? Changes in version 0.2 o added parnames, parnames<- o minor fix to allow "profiles" of 1-parameter models (skip fdHess call) o minor change to print method for mle results o tweaking "vecpar" (to allow parameter vectors in objective function) o removed fdHess/nlme dependency