The R Project SVN R-packages

Rev

Rev 8597 | Blame | Compare with Previous | Last modification | View Log | Download | RSS feed 

2026-07-15  Kurt Hornik  <Kurt.Hornik@wu.ac.at>

    * DESCRIPTION (Version, Date): 3.1-170.

2026-04-24  Sebastian Meyer  <Sebastian.Meyer@R-project.org>

    * DESCRIPTION (Suggests): remove SASmixed (archived).
    * inst/scripts/ch02.R: embed data(PBIB).

2026-03-27  Sebastian Meyer  <Sebastian.Meyer@R-project.org>

    * DESCRIPTION (Version, Date): 3.1-169

    * R/gnls.R: fix error message for missing start values
    when no "selfStart" model is used.

    * tests/gls.R: change gnls() start values for compatibility
    with OpenBLAS on aarch64.

2026-02-06  Sebastian Meyer  <Sebastian.Meyer@R-project.org>

    * tests/anova.gls.R: increase tolerance.

2025-03-31 Brian Ripley

    * DESCRIPTION: Version is 3.1-168
    * Update German translations

2025-01-27  Kurt Hornik  <Kurt.Hornik@wu.ac.at>

    * DESCRIPTION (Version): 3.1-167 (released)

2024-08-22  Sebastian Meyer  <Sebastian.Meyer@R-project.org>

    * R/corStruct.R (print.summary.corNatural):
    failed from wrong coef() method.

2024-08-21  Sebastian Meyer  <Sebastian.Meyer@R-project.org>

    * src/nlmefit.c (mixed_fcn): guard against non-finite results
    of 'internal_loglik', potentially causing infinite looping in
    'optif9'; reported for simulate.lme() in PR#17955 and PR#18433.
    * tests/predict.lme.R: add corresponding regression test.

2024-08-16  Sebastian Meyer  <Sebastian.Meyer@R-project.org>

    * tests/corFactor.R, tests/corMatrix.R: rename and add test of
    spatial correlation with groups.

    * src/corStruct.c (corStruct_recalc): fix iterating over 'Xy' to
    be compatible with spatial correlation structures, avoiding memory
    errors from nlme() (PR#18192) and gnls() (PR#17227).
    Thanks to Laurens Bogaardt for the useful reproducible example and to
    Mikko Korpela, Benjamin Tyner, and Ivan Krylov for initial debugging.
    * tests/nlme.R: add corresponding regression test.

2024-08-13  Kurt Hornik  <Kurt.Hornik@wu.ac.at>

    * DESCRIPTION (Version): 3.1-166 (released)

2024-08-12  Sebastian Meyer  <Sebastian.Meyer@R-project.org>

    * R/corStruct.R (Initialize.corARMA): catch unsupported values of
    'p' and 'q' (PR#18777, thanks to Liming Li).
    * tests/corStruct.R: added, with regression test.

2024-06-18  Sebastian Meyer  <Sebastian.Meyer@R-project.org>

    * tests/deviance.R, tests/nlme.R, tests/updateLme.R:
    use 'MASS' conditionally.

2024-06-16  Kurt Hornik  <Kurt.Hornik@wu.ac.at>

    * man/pairs.compareFits.Rd:
    * man/pairs.lmList.Rd:
    * man/pairs.lme.Rd:
    * man/plot.ACF.Rd:
    * man/plot.Variogram.Rd:
    * man/plot.augPred.Rd:
    * man/plot.gls.Rd:
    * man/plot.lmList.Rd:
    * man/plot.lme.Rd:
    * man/plot.nfnGroupedData.Rd:
    Add package anchors for links to targets from package 'lattice'.

2024-06-05  Sebastian Meyer  <Sebastian.Meyer@R-project.org>

    * DESCRIPTION (Depends): update to real minimum R version (R >= 3.6.0).

    * po, inst/po: update (installed) translations

    * man/simulate.lme.Rd, man/summary.lme.Rd: drop unneeded \donttest.

2024-04-29  Sebastian Meyer  <Sebastian.Meyer@R-project.org>

    * R/nlme.R (nlme.nlsList): fix scoping issue for nlsList() and
    self-starting nlme() calls with a preconstructed formula.

2024-04-28  Sebastian Meyer  <Sebastian.Meyer@R-project.org>

    * R/corStruct.R (corARMA): corARMA() no longer returns the
    dysfunctional/undocumented corIdent() structure, with a clear
    error message when neither 'p' nor 'q' are set > 0 (PR#17988).

2024-04-23  Sebastian Meyer  <Sebastian.Meyer@R-project.org>

    * R/VarCov.R (getVarCov.lme, getVarCov.gls): support
    'individual[s]' with only one observation (PR#16806).
    * tests/getVarCov.R: add corresponding regression tests.

2024-04-17  Sebastian Meyer  <Sebastian.Meyer@R-project.org>

    * R/lme.R (predict.lme): use formula() instead of
    re-evaluating the call (with the wrong scope), similar to PR#15892.

    * R/nlme.R (nlme.formula): nlme() finally works with fixed
    variance weights (PR#17712), thanks to the pointers by Bill
    Denney, Benjamin Tyner, and Ben Bolker.
    * tests/varFixed.R: added, based on BD's example.

    * R/varFunc.R (varIdent): warn about misguided varIdent(~1|g)

2023-12-08  Martin Maechler  <maechler@stat.math.ethz.ch>

    * tests/augPred_lab.R: manually do Hmisc  label(.) <- ..
    * DESCRIPTION (Version): 3.1-165; no longer suggesting 'Hmisc'.

2023-11-27  Martin Maechler  <maechler@stat.math.ethz.ch>

    * src/nlmefit.c: 2 x format 'int': s/%ld/%d/
    * DESCRIPTION (Version): 3.1-164

2023-07-31  Martin Maechler  <maechler@stat.math.ethz.ch>

    * R/nlme.R (nlme.formula, formula.nlme):
    * R/gnls.R (gnls, formula.gnls): store 'model' argument as 'formula'
    ==> trivial formula() methods, fixing PR#18559 by Ben Bolker.

2023-04-06  Martin Maechler  <maechler@stat.math.ethz.ch>

    * tests/sigma-fixed-etc.R (nlme): loosen tol for sM4$tTable[,"Std.Error"]
    * DESCRIPTION (Version): 3.1-163

2023-01-28  Martin Maechler  <maechler@stat.math.ethz.ch>

    * DESCRIPTION (Version): 3.1-162 (to be released)

    * R/corStruct.R (corIdent, *.corIdent): deprecated corIdent() and
    some of the "corIdent" methods.
    * man/nlme-deprecated.Rd: deprecation

2023-01-26  Sebastian Meyer  <Sebastian.Meyer@R-project.org>

    * src/matrix.h: fix return value in C declaration of Fortran
    subroutine 'RS'.  (PR#18438 by Lionel Henry.)

2023-01-23  Sebastian Meyer  <Sebastian.Meyer@R-project.org>

    * R/groupedData.R (nfGroupedData, nmGroupedData):
    fix deprecation logic to give a warning when not called
    internally.

    * R/groupedData.R (balancedGrouped): suppress spurious warning
    from testing for numeric row names.

2022-11-17  Martin Maechler <maechler@stat.math.ethz.ch>

    * DESCRIPTION (Version): 3.1-161 (released)

    * R/pdMat.R (str.pdMat): new method, fixing error for
       un-initialized, and misleading e.g., in `pdDiag` case.

2022-10-10  Brian Ripley

    * DESCRIPTION (Version): 3.1-160 (released)

2022-08-24  Brian Ripley

    * src/nlmefit.c: Fix for -Wstrict-prototypes.

2022-07-28  Martin Maechler  <maechler@stat.math.ethz.ch>

    * DESCRIPTION (Version): 3.1-159

    * R/gnls.R (gnls): 'if (deparse(params[[nm]][[3]]) != "1")' fails
    for long formulas (R-help list, thanks to Bill Dunlap, Jul 24)

2022-07-05  Brian Ripley

    * man/{gls,lme}: document the prohibition on  (rather than
    ignoring of) offset() terms.

2022-06-13  Brian Ripley

    * DESCRIPTION (Version): 3.1-158 (released)
    * R/lme.R (formula.lme): get the formula from the terms instead of
    re-evaluating the call (with the wrong scope), fixing PR#15892.
    * tests/scoping.R: add corresponding regression test.
    * R/nlsList.R: use inherits() instead of comparing class to string.

2022-03-25  Brian Ripley

    * DESCRIPTION (Version): 3.1-157 (released)

2022-03-24  Sebastian Meyer  <Sebastian.Meyer@R-project.org>

    * R/nlme.R (nlme.formula):
    * R/gnls.R (gnls): revert check for offset() terms in non-linear
    models (avoid symbolic interpretation of arithmetic expressions).
    * R/lme.R (formula.lme): defer fix for scoping issue to give more
    time to update reverse dependencies

2022-03-22  Brian Ripley

    * DESCRIPTION (Version): 3.1-156 (unreleased)

2022-03-18  Sebastian Meyer  <Sebastian.Meyer@R-project.org>

    * R/gls.R: keep "terms" from gls() for use in predict.gls(),
    fixing PR#18283.
    * R/zzMethods.R, NAMESPACE: drop formula-based terms.gls().
    * tests/gls.R: add corresponding regression test.

    * R/gls.R (formula.gls):
    * R/lme.R (formula.lme): get the formula from the terms instead of
    re-evaluating the call (with the wrong scope), fixing PR#15892.
    * tests/scoping.R: add corresponding regression test.

2022-03-17  Sebastian Meyer  <Sebastian.Meyer@R-project.org>

    * R/lme.R (lme.formula):
    * R/nlme.R (nlme.formula):
    * R/gls.R (gls):
    * R/gnls.R (gnls): stop() at offset() terms, fixing PR#17880.
    (They were previously ignored.)

2022-03-08  Sebastian Meyer  <Sebastian.Meyer@R-project.org>

    Improve handling of factor variables in predict() methods.

    * R/gls.R (predict.gls):
    * R/gnls.R (predict.gnls):
    * R/lme.R (predict.lme):
    * R/nlme.R (predict.nlme):
    call model.frame() with 'xlev' to make sure factor levels are
    compatible with the employed contrasts (fixing PR#17226),
    including to warn for numeric 'newdata' for factor variables,
    and to support character input (cf. PR#18312).
    * tests/gls.R:
    * tests/nlme2.R:
    * tests/predict.lme.R: add corresponding regression tests.

    * R/gnls.R (gnls): always return contrast _matrices_ not just names
    of contrast functions. Needed for the above fix and for
    consistency with the other modelling functions in nlme.

2022-03-05  Sebastian Meyer  <Sebastian.Meyer@R-project.org>

    * inst/CITATION: use citation(auto=meta) and add NLME book.

    * DESCRIPTION (Depends): finally require R >= 3.5.0.
    * R/lmList.R: drop old warnErrList() code, part of utils >= 3.5.0.

    * man/reStruct.Rd: late doc update for the pdClass default changed
    2000-07-07 (PR#17740).

2022-03-04  Sebastian Meyer  <Sebastian.Meyer@R-project.org>

    * R/lme.R (predict.lme): fix predictions for character-type newdata.
    The example in ch04.R produced wrong results with R >= 4.0.0 (PR#18312).
    * tests/predict.lme: add corresponding regression test.

    * R/newFunc.R (asOneFormula, getCovariateFormula, getResponseFormula):
    * R/newMethods.R (getGroupsFormula.default):
    fix environment of created formula (was local envir, now .GlobalEnv),
    in line with stats::asOneSidedFormula() in R 4.2.0.

    * inst/scripts/ch06.R: increase maxiter for nlsList() of Soybean data
    to only fail for a single plot, consistent with the book (p. 289).

2022-03-03  Sebastian Meyer  <Sebastian.Meyer@R-project.org>

    Setup extra tests conditional on nlme:::doExtras()

    * DESCRIPTION: add SASmixed to Suggests (used in code from NLME book)
    * inst/scripts: arrange for cleaner Rdiff() (drop proc.time(), tag
    platform-dependent output via ## IGNORE_RDIFF_*, reduce digits in tests)
    * tests/extras: added, including test scripts to run the full code
    from the installed book chapters (with reference output)
    * tests/nlme-stall.R: moved to extras

2022-03-02  Sebastian Meyer  <Sebastian.Meyer@R-project.org>

    * R/lmList.R (coef.lmList, summary.lmList): fix finding unique
    coefficient names when the set of estimated coefficients varies
    between subgroups (PR#16542); based on a patch by Ben Bolker.
    * tests/lmList.R: add corresponding regression test

2022-02-28  Sebastian Meyer  <Sebastian.Meyer@R-project.org>

    Clean up license information

    * DESCRIPTION (Authors@R): "R Core Team" rather than "R-core"
    * DESCRIPTION (License): dropped redundant "| file LICENCE"
    * LICENCE: removed (copy of GPL-2)
    * LICENSE.note: created from README
    * README: removed

2022-02-26  Sebastian Meyer  <Sebastian.Meyer@R-project.org>

    * R/VarCov.R (getVarCov.lme):
    * R/corStruct.R (Variogram.corSpatial):
    * R/lmList.R (predict.lmList):
    * R/lme.R (plot.ranef.lme): fix partial matching issues.

2022-02-25  Sebastian Meyer  <Sebastian.Meyer@R-project.org>

    Fix more scoping issues (detected by sourcing the installed
    chapter scripts in a local environment)

    * R/simulate.R (simulate.lme): evaluate constructed calls in the
    parent frame.
    * R/lme.R (augPred.lme): look for data in the parent frame.
    * R/gls.R (augPred.gls): (ditto)

2022-02-23  Sebastian Meyer  <Sebastian.Meyer@R-project.org>

    * R/simulate.R (simulate.lme): fix `useGen = TRUE` branch.

2022-01-13  Brian Ripley

    * DESCRIPTION (Version): 3.1-155 (released)
    * src: Use R_Calloc, R_Free.

2021-11-02  Sebastian Meyer  <Sebastian.Meyer@R-project.org>

    * R/nlme.R (nlme.nlsList, nlme.formula): fix more scoping issues
    with self-starting models by evaluating calls in the parent frame.
    * R/nlsList.R (nlsList.selfStart): (ditto)
    * tests/scoping.R: add corresponding regression tests.

2021-10-21  Sebastian Meyer  <Sebastian.Meyer@R-project.org>

    * R/gnls.R (gnls), tests/scoping.R: fix evaluation environment of
    the internal `nls` call used for self-starting models (PR#18157).

    * R/nlme.R (nlme.formula), tests/nlme.R: fix "unused argument" error
    when explicitly specifying `groups` in a self-starting model.

2021-09-23  Martin Maechler  <maechler@stat.math.ethz.ch>

    * DESCRIPTION (Version): 3.1-154 (unreleased)

    * R/gls.R (print.intervals.gls): do not print `attr(., "label")` twice;
    ditto in R/lme.R, PR#18196, with thanks to Johannes Ranke.

2021-09-07  Martin Maechler  <maechler@stat.math.ethz.ch>

    * R/groupedData.R ("[.groupedData"): change the default of `drop` to
    to FALSE for the case of `x[j]`, fixing PR#18177.

    * R/lme.R (print.summary.lme): change partial match to
    `x$coefficients`, PR#18184.

    * DESCRIPTION: Release 3.1-154

2021-08-05  Martin Maechler  <maechler@stat.math.ethz.ch>

    * R/zzMethods.R (.onLoad), R/VarCorr.R: remove code used only
    if(getRversion() < "3.3"); from Sebastian Meyer.

2021-03-09  Martin Maechler  <maechler@stat.math.ethz.ch>

    * R/varFunc.R (print.varFunc): use full 'unconstrained' argument
    name, fixing PR#17767, thanks to Sebastian Meyer.

    * DESCRIPTION (Version, Date): 3.1-153 (*not* yet for release)

2021-02-12 Brian Ripley

    * po: update fr translations.

2020-02-03  Brian Ripley

    * DESCRIPTION (Version, Date): 3.1-152 for 4.0.4.

2020-12-09  Martin Maechler  <maechler@stat.math.ethz.ch>

    * nlme/man/Remifentanil.Rd: change 2 URLs to 'moved to' locations.

2020-12-07  Martin Maechler  <maechler@stat.math.ethz.ch>

    * R/corStruct.R (print.corCompSymm, print.summary.corCompSymm): new,
    thanks to patch in PR#17990 by Sebastian Meyer; also fixing typo
    ("Uninitialized").

2020-10-29  Martin Maechler  <maechler@stat.math.ethz.ch>

    * DESCRIPTION (Version): 3.1-151
    * R/varFunc.R (varConstProp): constructor and methods, from
    PR#17954 by Johannes Ranke.

2020-10-02  Brian Ripley

    * DESCRIPTION (Version, Date): Prepare to release 3.1-150.

2020-09-24  Peter Dalgaard

        * R/VarCov.R tests/getVarCov.R: Patch set from Sebastian Meyer
          to fix group order issues PR#16744
        * tests/corFactor.R: missed commit from patch for PR#16110
        * INDEX: removed as it was out of date and autobuilt anyway

2020-08-21  Brian Ripley

    * R/pdMat.R: use parent.frame not sys.parent for data= args
    * DESCRIPTION (Version, Date): Prepare to release 3.1-149.

2020-08-20 Peter Dalgaard

    * NAMESPACE, R/corStruct.R, man/corFactor.corStruct.Rd,
      man/corMatrix.corStruct.Rd, tests/corFactor.R: Applied patch set
      from Sebastian Meyer to fix misnamed
      corFactor.compSymm -> corFactor.corCompSymm PR#16110

2020-07-20 Deepayan Sarkar

    * R/lme.R: Applied patch from Johannes Ranke to fix PR#17761

    * DESCRIPTION: update version / date (unreleased)

2020-05-13  Martin Maechler  <maechler@stat.math.ethz.ch>

    * DESCRIPTION (Contact, MailingList): R-help (*not* R-core)

2020-04-21  Martin Maechler  <maechler@stat.math.ethz.ch>

    * DESCRIPTION (Version): 3.1-148 (not yet released)

    * man/nlsList.Rd: fix stopifnot(.) check in examples.

2020-03-03  Brian Ripley

    * src/nlmefit.c: clean up code.

2020-03-03  Brian Ripley

    * DESCRIPTION (Version): 3.1-147 (released)

    * po: update, including updated German translations from Detlef Steuer

2020-03-11  Martin Maechler  <maechler@stat.math.ethz.ch>

    * DESCRIPTION (Version): 3.1-146 (unreleased)

    * man/Remifentanil.Rd: more information incl references.

2020-03-03  Brian Ripley

    * DESCRIPTION (Version): 3.1-145 (released)

    * src/nlmefit.c: avoid clang-UBSAN warning.

2020-02-06

    * DESCRIPTION (Version): 3.1-144 (released)

2020-02-02  Kurt Hornik  <Kurt.Hornik@wu.ac.at>

    * man/gnls.Rd: Drop docs for argument not in usage.

2019-12-10  Brian Ripley

    * DESCRIPTION (Version): 3.1-143 (released)

    * tests/augPred_lab.R: change yet another 'plucked from air' tolerance.

2019-11-07  Martin Maechler  <maechler@stat.math.ethz.ch>

    * DESCRIPTION (Version): 3.1-142 (released)

    * R/nlme.R: use vapply and is.matrix.
    * man/pairs.compareFits.Rd: change example.

2019-09-26  Brian Ripley

    * src/nlmefit.c: suppress warning from clang trunk.

2019-08-01  Kurt Hornik  <Kurt.Hornik@wu.ac.at>

    * DESCRIPTION (Version): 3.1-141 (released)

2019-07-27  Kurt Hornik  <Kurt.Hornik@wu.ac.at>

    * R/lme.R:
    * tests/nlme2.R:
    Robustify against using stringsAsFactors = FALSE by default.

2019-05-10  Brian Ripley

    * DESCRIPTION (Version): 3.1-140 (released)

2019-05-01  Brian Ripley

    * src/init.c: correct declaration of fit_nlme

2019-04-10  Brian Ripley

    * src/nlmefit.c: remove clang warning.

2019-04-08  Brian Ripley

    * DESCRIPTION (Version): 3.1-139 (released)

2019-04-08  Brian Ripley

    * R/groupedData.R: *GroupedData are exported, so use :: not :::

2019-04-07  Martin Maechler  <maechler@stat.math.ethz.ch>

    * Release 3.1-138.

2019-04-02  Brian Ripley

    * Start preparing to release 3.1-138 for R 3.6.0

2019-04-02  Martin Maechler  <maechler@stat.math.ethz.ch>

    * po/{R-de,de}.po: updated

2019-01-23  Martin Maechler  <maechler@stat.math.ethz.ch>

    * R/lme.R (lme.formula, lmeControl): new option
    'allow.n.lt.q=FALSE' by default now triggers error
    "fewer observations than random effects in all level <Q> groups".

    * src/nlmefit.c (finite_diff_Hess): prevent integer overflow (and
    later seg.fault) for large 'nTot' (already for npar >= 305).

2018-08-04  Martin Mächler  <maechler@stat.math.ethz.ch>

    * DESCRIPTION (Version): 3.1-138 (unreleased)

    * R/groupedData.R: deprecate the *undocumented* functions
    nmGroupedData() and nfGroupedData() in favor of their wrapper
    groupedData() (to allow subsequent code simplification).

2018-08-03  Martin Maechler  <maechler@stat.math.ethz.ch>

    * R/lmList.R (plot.intervals.lmList, ..), tests/lmList.R: make
    'xlab', 'ylab', 'strip' regular arguments (instead of wrongly
    trying to use them from `...`).
    * R/lme.R (plot.ranef.lme, ..): (ditto)

2018-05-08  Brian Ripley

    * tests/sigma-fixed-etc.R: increase tolerance for ATLAS run.

2018-04-07  Brian Ripley

    * DESCRIPTION (Version): 3.1-137 for release with R 3.5.0

    * po/*/.pot: updated

    * po/{R-de,de}.po: updated

2018-03-09  Brian Ripley

    * DESCRIPTION: Update Date, release for R-devel.

    * tests/nlme-stall.R: Add comments.

2018-02-27  Martin Mächler  <maechler@stat.math.ethz.ch>

    * R/nlme.R, man/nlmeControl.Rd: nlmeControl() gets new
    `msWarnNoConv` option. When it is true, as by default, nlme() now
    warns about non convergence in the LME step of the 'alternating
    algorithm'.  Applies e.g. also to the nlme-stall.R example.

2018-02-27  Brian Ripley

    * DESCRIPTION: Update Date. (Still unreleased.)
    Relax R (>= 3.5.0) to 3.4.0 so it can be released generally once
    3.5.0 is out.

    * src/init.c, src/nlmefit.h: Remove broken utilities added on
    2018-01-31.

    * src/pythag.c: New file, contents moved from init.c.

    * tests/nlme-stall.R: New regression test, optional.

    * tests/sigma-fixed-etc.R: Changed tolerance after win-builder
    run, change t7.fix.REML.lme to use optim() (as done for LME, and
    Solaris has false convergence with nlminb()).  Drop all uses of
    platform-dependent tolerances.

2018-02-26  Brian Ripley

    * src/init.c, src/rs.f: The 2018-01-31 change is broken (passes
    pointers as doubles).  Try a simpler approach, to replace PYTHAG
    in rs.f by a wrapped call to the C99 function hypot().

2018-02-16  Brian Ripley

    * DESCRIPTION (Version): 3.1-131.1, for R 3.4.4.
    (and 3.1-131) are available at
    https://svn.r-project.org/R-packages/branches/R-3-4-branch/nlme

    * tests/deparse.R: comment on failure with macOS's Accelerate.

2018-01-31  Martin Maechler  <maechler@stat.math.ethz.ch>

    * src/nlmefit.h, src/init.c (risnan, risfinite, ..): utilities to
    check for NaN, NA, Inf etc from Fortran.  [Since removed.]

    * src/rs.f (PYTHAG): prevent infinite loop checking for finite arg.
    as reported in https://stat.ethz.ch/pipermail/r-devel/2018-January/075459.html .

    * src/nlme.c (nlme_iterate): add R_CheckUserInterrupt()

2018-01-02  Martin Maechler  <maechler@stat.math.ethz.ch>

    * R/lmList.R (warnErrList): bug fix.
    Added to the 'utils' package for R-devel, keep here for R <= 3.4.x.

2018-02-16  Brian Ripley

    * DESCRIPTION (Version): 3.1-135.5 (released for R-devel only)

2018-01-23, 2018-02-16 Brian Ripley

    * tests/predict.lme.R: increase tolerance.

2018-01-01  Brian Ripley

    * DESCRIPTION (Version): 3.1-135 (released for R-devel only)
    * tests/anova.gls.R: increase tolerance.

2017-12-24  Brian Ripley

    * tests/anova.gls.R: increase tolerance.

2017-12-01  Brian Ripley

    * DESCRIPTION (Version): 3.1-134, R >= 3.5.0
    * R/VarCorr.R: correct breakage in r7326 (2017-03-07)
      (broke packages joineR and lme4).

2017-11-28  Brian Ripley

    * DESCRIPTION (Version): 3.1-133 (released for R-devel but withdrawn)
    * tests/coef.Rout.save: update for R-pre-3.5.0 changes to str()

2017-09-18  Martin Maechler

    * tests/nlme2.R: 'tweak and investigate platform dependent results'

2017-03-07  Martin Maechler

    * nlme/R/VarCorr.R, nlme/man/VarCorr.Rd: see below.
    * 'correct default for VarCorr.lme() to what it has really been.'
    * 'Fix print.*()'.
    * 'Cosmetic (not changing any results)'

2017-02-27  Martin Maechler

    * nlme/R/gnls.R, nlme/man/glsControl.Rd, nlme/man/gnlsControl.Rd,
      nlme/man/lmeControl.R:  'cosmetic'

2017-02-22  Martin Maechler

    * DESCRIPTION (Version): 3.1-132 (unreleased)
    * R/gnls.R, src/gnls.c: 'small cosmetic related to getting closer
    to the root cause of PR#17227'

2017-02-11 Brian Ripley

    * src/nlmefit.c: gcc -Wconversion pointed out misuse of long
    constants '0L' and  '1L'.

2017-02-06  Martin Maechler

    * DESCRIPTION (Version): 3.1-131, tweaked test tolerance to also
    pass openblas 0.2.18, in
    * tests/sigma-fixed-etc.R example 6

2017-01-23  Martin Maechler

    * DESCRIPTION (Version): 3.1-130
    * NAMESPACE, R/zzmethods.R, tests/deviance.R: workaround design
    error in 3.1-129 re deviance.

2017-01-17  Martin Maechler

    * R/groupedData.R (groupedData): must prepend 'nlme::' for call,
    fixing PR #17211.

2016-12-12  Martin Maechler

    * R/lme.R (Variogram.lme): fix PR #17192, with thanks to
    Sebastian Meyer.

2016-07-11 Martin Maechler

    * DESCRIPTION (Version): 3.1-129
    * NAMESPACE, R/zzMethods.R (confint, sigma, deviance) methods (the
    last giving an error).

2016-05-04 Martin Maechler

    * R/lme.R, etc (lmeControl): 'sigma = 0' is now allowed (and
    equivalent to 'sigma = NULL'); this fixes mgcv::gamm()'s use of
    the equivalent of 'do.call(lmeControl, lmeControl())'.

2016-04-26 Martin Maechler

    * DESCRIPTION (Version): 3.1-128
    * R/lme.R (qqnorm.lme): fixed bug introduced in v 3.1-127
    * R/lme.R, R/nlme.R, etc: "fixed sigma" now computes 'apVar'
    correctly; thanks to patches from Siem Heisterkamp
    * R/lme.R, R/gls.R: coef(summary(.)) now also works for "gls", "lme".
    * R/gls.R (anova.gls): fixed a very long standing bug in the F
    statistics for the ML case, thanks to Siem H.

2016-03-21 Martin Maechler

    * R/nlme.R (getParsNlme): wish of PR#16614 : now predict REs as NA
    for non-existing levels instead of signalling error.
    * R/*.R: cosmetics: More vapply(); quote(.) instead of as.name(),
    passing '...' down in print() methods, etc.

2016-03-16 Martin Maechler

    * DESCRIPTION (Version): 3.1-127
    * R/lmList.R, R/nlsList.R: same kind of error catching -> warning
    * */*.R: speedup cosmetics:  seq_along; vapply(); lengths();
    lapply(a, `[[`, i).  Less  eval(parse(text = ....))

2016-03-12 Martin Maechler

    * DESCRIPTION: version 3.1-126
    * R/nlme.R (predict.nlme): fix bug PR#16715; +cosmetic; also in
    * R/lmList.R, R/nlsList.R: better error catching -> warning()

2016-02-27 Brian Ripley

    * DESCRIPTION: version 3.1-125; released
    * tests/coef.Rout.save: update for R 3.3.x

2016-02-13 Brian Ripley

    * src/*.[ch]: remove use of S.h, clean up.
    * src/nls.c: Unused in R, so remove.

2016-01-25 Brian Ripley

    * src/*.c: include directly headers included via R.h.

2016-01-19 Martin Maechler

    * DESCRIPTION: version 3.1-124, and released

2016-01-19 Brian Ripley

    * src/{corStruct,gnls,nlme,nlmefit,nls}.c: Avoid clang warnings on
    implicit conversions.

2016-01-18 Martin Maechler

    * nlme/man/lmeControl.Rd, R/lme.R, etc: in lmeControl() etc,
    currently 'apVar' to FALSE when sigma is fixed, and document that
    somewhat.

2016-01-11  Martin Maechler

    * R/VarCorr.R, man/VarCorr.Rd: add optional '...' and remove 'rdig'
    from the argument list of the generic (for other methods).

2015-11-25 Bert van Willigen --- with modifications by MM
    * DESCRIPTION: version 3.1-123 (released 2016-01-17)
    * man/*Control.Rd: Fixed "sigma" (residual error) added.
    * R/gls.R, R/gnls.R, R/lme.R, R/nlme.R, R/reStruct.R: Fixed sigma added.
    * src/init.c, src/nlme.c, src/nlmefit.?:
    * tests/example_[1-9]_wequantify.R: 9 files added.
    * README: Credentials added.

2015-08-18 Luke Tierney
    * DESCRIPTION: version 3.1-122
    * src/nlmefit.c: Added PROTECT/UNPROTECT calls.

2015-06-29  Brian Ripley

    * DESCRIPTION: version is 3.1-121

    * NAMESPACE: tweak imports.

2015-02-20  Brian Ripley

    * DESCRIPTION: version is 3.1-120

2015-02-07  Brian Ripley

    * man/bdf.Rd: change markup for a dead URL

2015-01-29  Brian Ripley

    * po/*.de.po: updates.

2015-01-23  Brian Ripley

    * NAMESPACE: register lme method for simulate()

2015-01-10  Martin Maechler

    * remove non-useful legacy "code":  q(); emacs "Mode: S" etc

2015-01-09  Brian Ripley

    * DESCRIPTION: version is 3.1-119

    * man/bdf.Rd: update URL.

2014-10-20  Brian Ripley

    * R/*.R: remove unused assignments.

2014-10-07  Brian Ripley

    * DESCRIPTION: version is 3.1-118

2014-09-30 Brian Ripley

    * R/nlme.R: nlme with self-starting model incorrectly required 'start'.

    * tests/nlme2.R: regression test for above.

    * R/*, src/* correct copyright information.


2014-03-31  Brian Ripley

    * DESCRIPTION: version is 3.1-117

    * po/R-fr.po: update.

2014-03-28  Brian Ripley

    * DESCRIPTION: version is 3.1-116

    * R/{lme,newMethods,nlsList}.R: modify messages.
    * po/{R-pl,pl}.po: update.

2014-03-19  Brian Ripley

    * tests/missing.{R,Rout.save}: reduce 'digits' for cleaner results
    on platforms without extended-precision hardware.

    * po/R-de.po: Update from Detlef Steuer.

2014-03-09  Brian Ripley

    * DESCRIPTION: version is 3.1-115

    * R/{gls,gnls,lme,nmle}.R, man/{lmeScale,*Control}.Rd:
    remove msScale and nlmStepMax arguments which were used for nlm(),
    where that is not longer used.

2014-03-08  Brian Ripley

    * DESCRIPTION: version is 3.1-114

    * R/new-methods.R: labeling from fitted() was sometimes wrong as
    the 'groups' component lost row names. (PR#15678)

    * tests/fitted.R: add regression test.

    * R/lme.R: clean up

    * man/{lmeControl,lmeScale}.Rd: update..

2013-11-17  Brian Ripley

    * DESCRIPTION: version is 3.1-113

    * src/{matrix,nlOptimizer}.h: typos in guard names (clang 3.4)

2013-10-21  Martin Maechler

    * R/gls.R, man/gls-internal.Rd: remove redundant parentheses.

2013-10-03  Brian Ripley

    * man/{gls,lme}.Rd: add links, warn on no-covariate formulae.

2013-09-30, 2013-10-02  Brian Ripley

    * DESCRIPTION: version is 3.1-112

    * R/gls.R, man/glsControl.Rd, man/Initialize.glsStruct.Rd.
    Remove 'qrTol', which was unused and set to the non-existent
    .Machine$single.eps, that is to NULL.

    * NAMESPACE, man/gls-internal.Rd: export glsApVar and glsEstimate
    for use in package rms.

    * R/gls.R, lme.R: more use of integer constants.

    * man/*.Rd: classes in quotes, more links, remove CVS lines.

    * R/lme.R, R/newFunc.R, tests/predict.lme.R:
    fix asOneFormula and predict.lme for y ~ 0.
    (see https://stat.ethz.ch/pipermail/r-devel/2013-September/067600.html)

    * R/lme.R: Several of the return components were incorrect for a
    model of the form fixed = resp ~ 0.

2013-08-17  Brian Ripley

    * add start on ko translations.

2013-08-15  Brian Ripley

    * DESCRIPTION: version is 3.1-111
    * R/lmList.R: Adjust to cope with failed fits and fits with
    fewer coefficients returned by lm().  (Contributed by Ruth Ripley.)

2013-07-02  Brian Ripley

    * DESCRIPTION: version is 3.1-110,
    * R/{gls,gnls,lme}.R: Replace as.name() calls by quote(stats::nls) etc.
    * inst/LICENCE: remove.

2013-03-21  Brian Ripley

    * inst/CITATION: modernize.

2013-03-11  Brian Ripley

    * LICENCE: remove (it was not a licence file).

2013-01-26  Brian Ripley

    * DESCRIPTION: version is 3.1-109,
    depends on R (>= 3.0.0).

2013-01-26  Brian Ripley

    * DESCRIPTION: version is 3.1-108,
    depends on R (< 3.0.0) and has chol.f and rs.f in .Rbuildignore.

2013-01-23  Brian Ripley

    * man/[n]lme.Rd: add a note about the need to scale the response.
    * man/lmeControl.Rd: correct description of 'msTol' argument.
    * man/[n]lmeControl.Rd: correct description of 'opt' argument (it is
    a character string).
    * man/*.Rd: s/Jose/José/
    * R/[n]lme.R, man/[n]lmeControl.Rd: allow '...' for further
    control arguments to be based though to nlminb() and optim().


2013-01-13  Brian Ripley

    * DESCRIPTION: add accent on JCP's first name (and declare
    UTF-8 encoding). Add EISPACK authors.
    * src/rs.f: minimal version of eigen.f containing just RS and
    ancillaries (replacing eigen.f).

2013-01-12  Brian Ripley

    * DESCRIPTION: version is 3.1-107
    * src/{chol,eigen}.f: port files from R (where they are
      deprecated/defunct).

2012-12-06  Kurt Hornik

    * DESCRIPTION: version is 3.1-106, add roles to Authors@R.

2012-11-28  Brian Ripley <Brian.Ripley@R-project.org>

    * NAMESPACE: import selectively from lattice and stats.

2012-09-24  Brian Ripley <Brian.Ripley@R-project.org>

    * DESCRIPTION: version is 3.1-105
    * R: improve messages for translation.

2012-09-01  Brian Ripley <Brian.Ripley@R-project.org>

    * po: add Polish translations.

2012-08-30  Brian Ripley <Brian.Ripley@R-project.org>

    * src/init.c: force use of registered symbols in R >= 2.16.0

2012-05-23  Brian Ripley <Brian.Ripley@R-project.org>

    * DESCRIPTION: version is 3.1-104
    * src/gnls.c, src/nlme.c, src/nlmefit.[ch]:
    type mismatches shown by clang's -Wconversion

2012-04-25  Brian Ripley <Brian.Ripley@R-project.org>

    * src/nlmefit.c: type mismatch shown by clang's -Wconversion

2012-03-29  Brian Ripley <Brian.Ripley@R-project.org>

    * src/nlOptimizer.c: increase PROTECTion (reported by Andrew Runnalls).

2012-02-15  Brian Ripley <Brian.Ripley@R-project.org>

    * inst/CITATION: protect against TRE bug in UTF-8 locales.

2012-01-14  Brian Ripley <Brian.Ripley@R-project.org>

    * DESCRIPTION: version is 3.1-103

2011-12-25  Brian Ripley <Brian.Ripley@R-project.org>

    * DESCRIPTION: needs Suggests: MASS for data(package=)

2011-07-19  Brian Ripley <Brian.Ripley@R-project.org>

    * DESCRIPTION: version is 3.1-102
    * R/newMethods.R: getData.nls (aka getData.lme) used the na.action
    of the call on the whole data and not on the model frame: should
    have used the na.action component of the result.
    * tests/{lme.R,lme.Rout.save,ss2.rda}: test for above.
    * man/simulate.lme: document as method.

2011-07-11  Brian Ripley <Brian.Ripley@R-project.org>

    * COPYING: rename to GPL-2
    * LICENCE, inst/LICENCE: adjust wording accordingly

2011-05-30  Brian Ripley <Brian.Ripley@R-project.org>

    * R/newFunc.R: name arguments to lapply() when ... is present.

2011-05-06  Brian Ripley <Brian.Ripley@R-project.org>

    * DESCRIPTION: version is 3.1-101

2011-04-23  Martin Maechler  <maechler@stat.math.ethz.ch>
2011-04-14  Martin Maechler  <maechler@stat.math.ethz.ch>

    * man/augPred.Rd, R/groupedData.R, R/gls.R, R/lme.R, R/simulate.R
    Changes he failed to document.

2011-03-26  Brian Ripley <Brian.Ripley@R-project.org>

    * src/{nlmefit,pdMat}.c: remove unused variables detected by
    -Wunused-but-set-variable

2011-03-11  Brian Ripley <Brian.Ripley@R-project.org>

    * Release 3.1-100 for R 2.13.0

    * NAMESPACE: no longer re-export BIC

    * man/BIC.Rd: removed.

2011-02-23  Brian Ripley <Brian.Ripley@R-project.org>

    * DESCRIPTION: version is 3.1-100, require R >= 2.13

    * man/BIC.Rd: make a stub referring to stats::BIC()

    * Everywhere: remove obsolete email addresses of former authors

2011-02-23  Brian Ripley <Brian.Ripley@R-project.org>

    * DESCRIPTION: version is 3.1-99 (unreleased, used for R-devel),
    remove email addresses of former authors (and JCP has moved).

    * R/{new,zz}Methods.R:
    Put all BIC() methods in one place, as the default method.

    * man/BIC.Rd: remove aliases to methods (not all of which actually
    existed)

    * man/BIC.logLik.Rd: merge into BIC.Rd

    * tests/lme.Rout.save: update for R 2.13.0.

    * NAMESPACE, R/newMethods.R: prepare for migration of BIC() to stats.

    * R/{gls,gnls,lme}.R, NAMESPACE: add nobs() methods.

2011-02-10  Brian Ripley <Brian.Ripley@R-project.org>

    * DESCRIPTION (Version, Date): 3.1-98

2011-02-04  Brian Ripley <Brian.Ripley@R-project.org>

    * data/{MathAchieve,bdf}.rda: resave with xz compression.
    * DESCRIPTION: depend on R (>= 2.10).

2010-10-01  Brian Ripley <Brian.Ripley@R-project.org>

    * DESCRIPTION (Version, Date): 3.1-97

2010-09-22  Brian Ripley <Brian.Ripley@R-project.org>

    * man/residuals.gls.Rd: incorrect default, see
    https://stat.ethz.ch/pipermail/r-devel/2010-September/058487.html

2009-12-22  Brian Ripley <Brian.Ripley@R-project.org>

    * man/ACF.gls.Rd: has \seealso to itself: seems ACF.lme was meant.

2009-12-18  Brian Ripley <Brian.Ripley@R-project.org>

    * DESCRIPTION: add BugReports field.

2009-10-12,16  Brian Ripley <Brian.Ripley@R-project.org>

    * DESCRIPTION (Version, Date): 3.1-96
    * po/R-de.po, po/de.po: update

2009-09-16  Martin Maechler  <maechler@stat.math.ethz.ch>

    * DESCRIPTION (Version, Date): 3.1-95
    * R/nlme.R (nonlinModel): use unname() instead of clearNames()

2009-09-02  Brian Ripley <Brian.Ripley@R-project.org>

    * DESCRIPTION (Version, Date): 3.1-94
    * man/*.Rd: remove unnecessarily anchored links, including several
    incorrect ones.

2009-08-16  Brian Ripley <Brian.Ripley@R-project.org>

    * DESCRIPTION (Version, Date): 3.1-93
    * R/lmList.R, tests/lmList.R: bug fix for predict.lmList (PR#13788)
    * R/gls.R, tests/anova.gls.R: bug fix for anova.gls (PR#13567)

2009-05-30  Brian Ripley <Brian.Ripley@R-project.org>

    * R/gnls.R, R/corStruct.Rd: minor message wording changes from
    Debian 'bug' report.
    * R/R-de.po: add

    [ Missing entries for 3.1-9[12] which were concerned with German
    translations. ]

2009-01-16  Brian Ripley <Brian.Ripley@R-project.org>

    * man/*.Rd: expand tabs

2008-12-30  Brian Ripley <Brian.Ripley@R-project.org>

    * man/summary.pdMat.Rd: avoid embedded newlines in \code strings.

2008-12-27  Brian Ripley <Brian.Ripley@R-project.org>

    * DESCRIPTION (Version, Date): 3.1-90
    * R/gls.R: call dist correctly (PR#13418)

2008-12-19  Brian Ripley <Brian.Ripley@R-project.org>

    * man/*.Rd: clean up some reference formats and misuse of $...$.

2008-12-11  Brian Ripley <Brian.Ripley@R-project.org>

    * man/*.Rd: spelling
    * src/nlOptimizer.c: missing PROTECT

2008-06-07  Brian Ripley <Brian.Ripley@R-project.org>

    * DESCRIPTION (Version, Date): 3.1-89
    * R/lme.R: fitted.lme() always dropped, when it should drop for
    one column but not for one row.
    * tests/lme.R, lme.Rout.save: add regression test for
    single-row predictions.

2008-04-10  Brian Ripley <Brian.Ripley@R-project.org>

    * R/zzMethods.R: add .onUnload() to unload the DLL.

2008-03-30  Brian Ripley <Brian.Ripley@R-project.org>

    * DESCRIPTION (Version, Date): 3.1-88

2008-02-19  Brian Ripley <Brian.Ripley@R-project.org>

    * man/logLik.lme.Rd, man/logLik.gnls.Rd: correct description of
    'REML' argument.

2008-02-19  Brian Ripley <Brian.Ripley@R-project.org>

    * inst/CITATION: add file (essentially what citation() would
    produce, without the warning).

2008-02-18  Brian Ripley <Brian.Ripley@R-project.org>

    * DESCRIPTION (Version, Date): 3.1-87
    * R/gnls.R, tests/gls.R: fix for PR#10364, missing row names
    on predict.gnls.

2008-02-11  Duncan Murdoch <murdoch@stats.uwo.ca>

    * R/*.R: Return invisible(x) from numerous print methods.

2007-12-30  Douglas Bates <bates@stat.wisc.edu>

    * [r4884] man/plot.nffGroupedData.Rd: Correct malformed link reported
      by Yihui Xie <xieyihui@gmail.com>. Remove reference to nonexistent
      URL.
    * [r4885] Initialize.reStruct.Rd, collapse.groupedData.Rd,
      fitted.lmeStruct.Rd, lme.Rd, lme.groupedData.Rd, lme.lmList.Rd,
      logLik.reStruct.Rd, nlme.nlsList.Rd, plot.nfnGroupedData.Rd,
      plot.nmGroupedData.Rd, recalc.reStruct.Rd, residuals.lme.Rd,
      residuals.lmeStruct.Rd: Remove reference to nonexistent URL

2007-10-04  Kurt Hornik

    * DESCRIPTION (Version, Date): 3.1-86

2007-10-04  Brian Ripley <Brian.Ripley@R-project.org>

    * R/corStruct.R: check for uninitialized corARMA structures.

2007-09-17  Brian Ripley <Brian.Ripley@R-project.org>

    * DESCRIPTION (Version, Date): 3.1-85
    Also, use approved form of GPL (>=2).
    * R/varFunc.R: fix for PR#9831.
    * tests/coef.R: test for the above.
    * tests/lme.Rout.save: update format for R >= 2.6.0.
    * man/corRatio.Rd: wrong formula, reported to R-devel by
    Ben Bolker on 2007-09-03.

2007-07-26  Brian Ripley <Brian.Ripley@R-project.org>

    * DESCRIPTION (Version, Date): 3.1-84
    * LICENCE, inst/LICENCE: clarify this is licensed under GPL >= 2.
    * README, R/*.R, src/*.[ch]: Update source for GPL.
    * man/*.Rd: add header with licence conditions

2007-07-02  Brian Ripley <Brian.Ripley@R-project.org>

    * R/varFunc.R: varIdent did not look at its 'fixed' arg, PR#9765
    * tests/varIndent.R: test for the above.

2007-06-13  Brian Ripley <Brian.Ripley@R-project.org>

    * R/VarCov.R: correct getVarCov.gls, PR#9752

2007-06-13  Brian Ripley <Brian.Ripley@R-project.org>

    * DESCRIPTION (Version, Date): 3.1-83
    Also, update authors and email addresses where known.
    * R/lme.R: called quantile on potentially missing values.
    * tests/missing.{R,Rout.save}: test summary methods

2007-06-12  Brian Ripley <Brian.Ripley@R-project.org>

    * DESCRIPTION (Version, Date): 3.1-82
    Also, R >= 2.4.0 because of seq_along.
    * tests/lme.Rout.save: back to 2.5.1 layout.

2007-05-29  Brian Ripley <Brian.Ripley@R-project.org>

    * R/{gls,lme,pdMat}.R: qualify uses of nchar() by 'type'.
    * R/lme.R: layout of over-long error message.
    * tests/lme.Rout.save: update for above and for 2.6.0 layout.
    * R/{gls,gnls,lme,nlme}.R: copy row names to fitted values and
    residuals.
    * R/{gls,gnls,lme}.R: support naresid and napredict.
    * man/{residuals,fitted}.lme.Rd: document this.
    * tests/missing.{R,Rout.save}: tests of missing-value support.

2007-05-20,24  Brian Ripley <Brian.Ripley@R-project.org>

    * R/*.R: partial matching issues, switch to seq_along.

2007-05-15  Brian Ripley <Brian.Ripley@R-project.org>

    * DESCRIPTION (Version, Date): 3.1-81
    * NAMESPACE, R/zzMethods.R: migrate vcov methods from stats to here.

2007-05-04  Brian Ripley <Brian.Ripley@R-project.org>

    * man/{Extracr.pdMat,getdata.*,summary.varFunc}.Rd: remove
    surplus }.

2007-03-30  Brian Ripley <Brian.Ripley@R-project.org>

    * DESCRIPTION (Version, Date): 3.1-80
    * R/nlme.R (contrMat): Allow for expressions in 'fixed'.
    * tests/contrMat.R: regression test for above.
    * R/gls.R: bug fix for the case of y ~ 0.
    * tests/gls.R: regression test for above.
    * R/{lme,lmList,newMethods}.R: patch for lattice 0.15-x.

2007-01-30  Brian Ripley <Brian.Ripley@R-project.org>

    * DESCRIPTION (Title): Capitalize for consistency.

2006-12-29  Brian Ripley <Brian.Ripley@R-project.org>

    * DESCRIPTION (Version, Date): 3.1-79

    * R/corStruct.R, R/pdMat.R, R/reStruct.R, man/corStruct.Rd,
    man/pdMat.Rd, man/reStruct.Rd: add '...' argument to as.matrix()
    methods. (Needed for 2.5.0, but works in earier versions.)

2006-12-10  Douglas Bates  <bates@stat.wisc.edu>

    * DESCRIPTION (Version, Date): 3.1-78

2006-12-06  Douglas Bates  <bates@stat.wisc.edu>

    * R/nlme.R (contrMat): Ensure that the contrasts component is a
    list of matrices by expanding function names if necessary.

2006-11-29  Douglas Bates  <bates@stat.wisc.edu>

    * DESCRIPTION: Prepare for new version.
    * R/newMethods.R (getData.nls): getData.nls for R's nls objects
    * inst/scripts/ch*.R: update Pinheiro and Bates (2000) scripts

2006-11-29  Brian Ripley <Brian.Ripley@R-project.org>

    * R/reStruct.R: fix PR#9262, multiple-line deparse.
    * R/gls.R, groupedData.R, lme.R, lmList.R, newMethods,R, nlme.R:
    more deparse width issues.

2006-09-28  Douglas Bates  <bates@stat.wisc.edu>

    * R/nlme.R (getParsNlme): backed out change that caused failure
    reported by Christian Ritz.

2006-09-18  Douglas Bates  <bates@stat.wisc.edu>

    * R/VarCorr.R (VarCorr): Clean up problems in this file and
    several others detected by codetools.

2006-09-11  Douglas Bates  <bates@R-project.org>

    * R/nlme.R (getParsNlme): '[[' not '[' in case start$fixed is a list
    * R/gls.R,gnls.R,newMethods.R,nlme.R,pdMat,Rsimulate.R: fix
    problems exposed by codetools.

2006-07-31  Douglas Bates  <Douglas.Bates@R-project.org>

    * man/ranef.lme.Rd: Modified the description of the `standard'
    argument to more clearly describe its effect.

2006-07-24  Douglas Bates  <Douglas.Bates@R-project.org>

    * R/lme.R: Patch by Andrew Robinson <A.Robinson@ms.unimelb.edu.au>
    and Berwin Turlach <berwin@maths.uwa.edu.au> to set maximum number
    of function calls in nlminb.

2006-07-03  Douglas Bates  <Douglas.Bates@R-project.org>

    * DESCRIPTION: Version is 3.1-75

    * man/{many files}: Documentation changes contributed by Spencer
    Graves <spencer.graves@pdf.com>.  Most of the changes add Pinheiro
    and Bates (2000) as a reference or add additional \seealso{}
    entries.  Some examples from Pinheiro and Bates (2000) have also
    been added.

    * R/nlme.R: reformat the code, typo.

2006-07-03  Martin Maechler  <maechler@stat.math.ethz.ch>

    * R/lme.R (lme.formula): when control()$returnObject is TRUE, only
    warn instead of stop() in the case of non-convergence.
    * tests/lme.R: test the above
    * tests/lme.Rout.save:

2006-06-02 Brian Ripley <Brian.Ripley@R-project.org>

    * DESCRIPTION: Version is 3.1-74

2006-05-31 Brian Ripley <Brian.Ripley@R-project.org>

    * R/lme.R: Keep terms and optional model frame in lme.formula.
    Use terms for prediction (get safe prediction that way).
    Add keep.data argument to lme and methods.
    * man/lme*.Rd: update for keep.data argument.
    * R/newMethods.R: make use of saved data in getData methods.
    * tests/getData.R, tests/predict.lme.R: new tests.

2006-05-24 Brian Ripley <Brian.Ripley@R-project.org>

    * R/gls.R, man/comparePred.Rd:  'minimum' and 'maximum' were
    evaluated on a formula in the comparePred methods.  Instead,
    evaluate 'primary' in the 'data' used to fit object1, and document
    what is done.  (Reported by Mark Difford.)

2006-05-22 Brian Ripley <Brian.Ripley@R-project.org>

    * DESCRIPTION: Version is 3.1-73

2006-05-21 Brian Ripley <Brian.Ripley@R-project.org>

    * LICENCE, inst/LICENCE: add and install a licence file.

2006-05-18 Brian Ripley  <Brian.Ripley@r-project.org>

    * R/groupedData.R: adjust method of as.data.frame.
    * R/newFunc.R: fix bug in gsummary with objects that have
    classes, e.g. Hmisc labels.  (Reported by Mark Difford, in a
    private followup to
    https://stat.ethz.ch/pipermail/r-help/2006-May/094549.html.)
    * tests/augPred_lab.R: tests for the above fix.

2006-05-12 Brian Ripley  <Brian.Ripley@r-project.org>

    * R/gls.R, R/gnls.R, R/nlme.R: missing data=data in terms calls,
    needed to resolve '.' on rhs of formulae.

    * DESCRIPTION: Version is 3.1-72

2006-03-29 Brian Ripley  <Brian.Ripley@r-project.org>

    * R/newMethods.R: as.list() did not work on symbols in R.

2006-03-14 Brian Ripley  <Brian.Ripley@r-project.org>

    * R/lme.R, man/Variogram.Rd: make example work (na.omit in
    the wrong place).

2006-03-10 Brian Ripley  <Brian.Ripley@r-project.org>

    * DESCRIPTION: Version is 3.1-71

2006-02-17 Brian Ripley  <Brian.Ripley@r-project.org>

    * R/nlsList.R: need to allow for multi-line deparse.
    * tests/deparse.R: add regression test

2006-01-05  Brian Ripley  <Brian.Ripley@r-project.org>

    * DESCRIPTION: Version is 3.1-70
    * src/init.c, src/Makevars: make use of gcc4 visibility where supported
    * inst/mlbook/ch0[45].R: remove unnecessary use of data()
    * inst/scripts/ch0[58].R: all allowed versions of R have nlminb
    * R/gls.R: element 'optimMethod' in the return value for
    glsControl was unnamed.
    * R/{gls,ngls,lme,nlme}.R, man/{gls,ngls,lme,nlme}Control.Rd:
    let the user (rather than the version of R) choose the optimizer
    to be used.

2006-01-05  Brian Ripley  <Brian.Ripley@r-project.org>

    * src/init.c: missing registration for ARMA_constCoef
    * inst/scripts/ch02.R: correct argument names in simulate.lme call

2005-12-31  Brian Ripley  <Brian.Ripley@r-project.org>

    * DESCRIPTION: Version is 3.1-69, require 2.3.0.
    * NAMESPACE: use registration.
    * R/*.R: use registered symbols.

2005-12-31  Brian Ripley  <Brian.Ripley@r-project.org>

    * DESCRIPTION: Version is 3.1-68.
    * src/init.c: new file to register symbols
    * R/corStruct.R: corrected one use of PACKAGE="base"
    * R/lmList.R: remove attempted use of 'frame = 1'

2005-12-09  Deepayan Sarkar  <deepayan@stat.wisc.edu>

    * R/newMethods.R, R/groupedData.R: fix for problem described in
    https://stat.ethz.ch/pipermail/r-help/2005-December/082890.html
    (superpose.line/symbol components were being truncated to length
    of an arbitrary component).


2005-12-09  Brian Ripley  <Brian.Ripley@r-project.org>

    * DESCRIPTION: Version is 3.1-67.
    * R/gls.R: Replace do.call by eval.parent on massaged call.
    * po/fr.R, po/R-fr.R, inst/po: add French translation by
    Philippe Grosjean and automated en@quot translation.

2005-08-25  Douglas Bates  <bates@stat.wisc.edu>

    * R/gls.R: Added option "optimMethod" to gls, gnls, and lme.

2005-08-24  Douglas Bates  <bates@wisc.edu>

    * R/lme.R (print.intervals.lme): Replace calls to print.matrix by
    print(as.matrix()).

2005-07-25  Douglas Bates  <bates@bates1-home>

    * inst/scripts/ch08.R: Cleaned up examples in ch04, ch06, ch08.

    * src/matrix.h: Include R_Ext/Linpack.h in addition to R_Ext/Applic.h

2005-07-19  Douglas Bates  <bates@stat.wisc.edu>

    * R/gnls.R: Added definition of convIter when using nlminb or when
    using optim.  Reindented this and other R source files.

2005-07-12  Douglas Bates  <bates@bates1-home>

    * inst/scripts/ch01.R: Clean up and check these scripts

    * src/nlmefit.c (gls_loglik): re-write contributed by ?? (I
    accidentally deleted the email message).

    * R/lme.R (lme.formula): Use nlminb if available.


2005-06-08  Douglas Bates  <bates@bates1-home>

    * R/lme.R (predict.lme): Fixed bug when using a name for a formula
    in original fit. (Reported by Arne.Mueller@sanofi-aventis.com)

2005-05-20  Douglas Bates  <bates@bates1-home>

    * man/contr.SAS.Rd: Removed

2005-04-04  Douglas Bates  <bates@bates1-home>

    * R/newFunc.R: Removed contr.SAS definition as it has been revised
    and moved into the stats package for R-2.1.0

2005-03-02  Douglas Bates  <bates@bates1-home>

    * src/nlmefit.c (gls_loglik): Replaced rnkm1 by p.

    * po/,inst/po/: Added localization support.

2004-11-04  Douglas Bates  <bates@bates1-home>

    * src/nlmefit.c (dimFree): Fix memory leak reported by Brian Ripley.

2004-11-03  Douglas Bates  <bates@bates1-home>

    * man/varWeights.Rd: Remove all calls to data()

    * man/corExp.Rd: Remove all references to the mva package

2004-08-10  Douglas Bates  <bates@bates1-home>

    * man/pairs.compareFits.Rd: fix syntax errors (new check available
    in R CMD check)

2004-01-14  Douglas Bates  <bates@stat.wisc.edu>

    * man/getCovariate.varFunc.Rd: Syntax correction from Kurt.

2003-12-16  Douglas Bates  <bates@stat.wisc.edu>

    * install.R: Create stored image

2003-08-29  Douglas Bates  <bates@stat.wisc.edu>

    * tests/lme.Rout.save (Coefficients): Now consistent with
    i386-linux result.

    * src/base.h: Replace R_ext/Mathlib.h by Rmath.h

2003-08-09  Douglas Bates  <bates@stat.wisc.edu>

    * man/Covariate.varFunc.Rd, man/Matrix.pdMat.Rd, man/Matrix.reStruct.Rd, man/Names.pdMat.Rd, man/Names.reStruct.Rd, man/coef.corStruct.Rd, man/coef.modelStruct.Rd, man/coef.pdMat.Rd, man/coef.reStruct.Rd, man/coef.varFunc.Rd, man/getData.lmList.Rd, man/gnls.Rd:
    Use \method for replacement method documentation as per Rd changes

    * R/lme.R:
    Changed check for intercept-like column in getFixDF (PR#2418)

2003-06-12  Douglas Bates  <bates@stat.wisc.edu>

    * DESCRIPTION: New release

    * man/ranef.lmList.Rd, man/nlme.Rd: Overdocumented arguments

2003-06-10  Douglas Bates  <bates@stat.wisc.edu>

    * man/MathAchSchool.Rd, man/MathAchieve.Rd, ChangeLog, DESCRIPTION, R/groupedData.R, inst/scripts/ch03.R, inst/scripts/ch04.R:
    Release 3.1-42

    * inst/scripts/ch01.R: Protecting power in formula

2003-06-10  Douglas Bates  <bates@stat.wisc.edu>

    * R/groupedData.R: Fixed bug in groupedData constructor when outer
    is not missing.

2003-05-16  Douglas Bates  <bates@stat.wisc.edu>

    * src/nlmefit.c (internal_EM): Fix by Peter Dalgaard for compound
    symmetry EM calculations (PR#2985)

2003-05-05  Douglas Bates  <bates@stat.wisc.edu>

    * R/lmList.R (update.lmList): Changed this and several other
    update methods.

    * R/groupedData.R: Use eval instead of do.call (Dieter Menne)

2003-04-17  Douglas Bates  <bates@stat.wisc.edu>

    * NAMESPACE: contributed by Brian Ripley

    * man/print.corStruct.Rd: Removed this and many other redundant
    documentation files for S3 methods.

    * R/<many files> Remove redundant GPL notices.

2003-03-25  Douglas Bates  <bates@stat.wisc.edu>

    * R/nlme.R (nonlinModel): Moved definition from AAA.R to here

    * man/summary.nlsList.Rd: Remove references to summary.nls

    * R/newMethods.R: Remove vestigial definitions of AIC.lm

    * R/zzMethods.R: Remove references to AIC.lm

2003-02-08  Douglas Bates  <bates@stat.wisc.edu>

    * R/lme.R (lme.formula): Removed hessian=TRUE in calls to optim as
    suggested by Brian Ripley.

2003-01-30  Douglas Bates  <bates@stat.wisc.edu>

    * R/lme.R:
    Incorporate change from Brian Ripley in response to Dieter Menne

    * man/lme.lmList.Rd, man/lme.groupedData.Rd, man/lme.Rd:
    Documentation of the contrasts argument added (Greg Warnes).

2003-01-30  Douglas Bates  <bates@stat.wisc.edu>

    * R/lme.R (lme.formula): Incorporate Greg Warnes' changes for
    the contrasts argument.

    * R/nlsList.R: Correct check for is.null(class(FUN))

2002-12-27  Douglas Bates  <bates@stat.wisc.edu>

    * R/newFunc.R (gsummary): Fix comparison in check for varying
    columns (PR#1775 - bug report with fix provided by Dieter Menne.

2002-12-11  Douglas Bates  <bates@stat.wisc.edu>

    * R/lme.R (print.lme): Removed one more as.vector()

    * R/varFunc.R (print.summary.varFixed): changed  instances of
    deparse(as.vector(x)) to deparse(x) here and in many other places.

2002-09-10  Saikat DebRoy

    * nlsList.R (nlsList.formula): Use "try" instead of "restart".

2002-08-09  Douglas Bates  <bates@stat.wisc.edu>

    * R/groupedData.R: Fixed creation of grouped data from a formula
    only

    * R/newGenerics.R: Renamed the initialize generic and all methods
    and calls to Initialize so as not to conflict with the methods
    package.

2002-07-23  Douglas Bates  <bates@stat.wisc.edu>

    * R/lme.R (update.lme): Fix formula handling in update methods
    (Brian Ripley).

2002-04-20  Douglas Bates  <bates@stat.wisc.edu>

    * R/nlme.R: Bug fix by SDR - convergence criterion was being
    calculated incorrectly.

2002-04-04  Douglas Bates  <bates@stat.wisc.edu>

    * man/getVarCov.Rd: Added an example of the type="marginal"
    Corrected misplaced brace.

    * R/VarCov.R (getVarCov.lme): Added the getVarCov functions
    contributed by Mary Lindstrom <lindstro@biostat.wisc.edu>

2002-03-27  Douglas Bates  <bates@stat.wisc.edu>

    * R/lme.R (anova.lme): Change strvar != "NA" to !is.na(strvar)

2002-03-05  Douglas Bates  <bates@stat.wisc.edu>

    * man/ACF.gls.Rd: Changed Jose's email address on this and all the
    other man pages.

    * R/varFunc.R: Incorporate fix from Thomas Lumley and from Setzer
    Woodrow for the logic where ifelse(length(el) == 1, is used
    instead of if(length(el) == 1)

    * R/pdMat.R: Change all vestigial instances of T and F to TRUE and
    FALSE.

2001-12-18  Douglas Bates  <bates@stat.wisc.edu>

    * R/corStruct.R: Finally resolved the print.summary.corSymm
    controversy so it prints and it prints the correct value.

2001-12-10  Douglas Bates  <bates@stat.wisc.edu>

    * R/lme.R: unclassed the logLik component of an lme summary before
    trying to put it into a data frame for printing.

2001-12-01  Douglas Bates  <bates@stat.wisc.edu>

    * R/zzMethods.R: Removed the .Alias calls.

2001-10-30  Douglas Bates  <bates@stat.wisc.edu>

    * man/pairs.lme.Rd: Modified many documentation files to pass the
    codoc checks.

2001-10-23  Douglas Bates  <bates@stat.wisc.edu>

    * man/ACF.lme.Rd: Modified many, many method definitions and
    documentation files to get them to pass the R-1.4.0 (in
    development) package checking process.

2001-10-08  Douglas Bates  <bates@stat.wisc.edu>

    * R/lmList.R: Fixed definition of logLik.lmList to work with R
    (but it is still rather ugly).

    * R/varFunc.R: Changed definition of logLik.varComb to return a
    logLik object.

2001-10-05  Douglas Bates  <bates@stat.wisc.edu>

    * R/lmList.R: Fixed selection of id[subscripts] in the panel
    function of pairs.lmList (and several other functions) so it does
    not produce a warning from evaluating is.na(NULL).

2001-09-17  Douglas Bates  <bates@stat.wisc.edu>

    * R/nlme.R: Incorporated a change from
    Setzer.Woodrow@epamail.epa.gov for using nested grouping factors
    in nlme.

2001-06-20  Douglas Bates  <bates@stat.wisc.edu>

    * R/varFunc.R: Changes in calling sequences and in documentation
    so the package passes codoc() checks.  These affect many files in
    both the R and the man directory.  No changes of substance.

2001-05-29  Martin Maechler <maechler@stat.math.ethz.ch>

    * R/zzMethods.R: Use .Alias()

    * R/newGenerics.R:  Generic AIC()
    * R/newMethods.R:   logLik.lm()   moved to "base R"
    * man/AIC.Rd,
    * man/AIC.logLik.Rd: AIC.logLik() and AIC.lm() as well
    * man/logLik.lm.Rd:

2001-03-28  Douglas Bates  <bates@stat.wisc.edu>

    * R/simulate.R: Modified simulate.lme for default pdMat class of
    pdLogChol.

2001-02-24  Douglas Bates  <bates@stat.wisc.edu>

    * R/newFunc.R: Removed the old fake xyplot function.  The new work
    by Murrell and Sarkar will make this superfluous.

    * man/MEestimate.Rd: Added manual pages for MEEM, MEdecomp,
    MEestimate.

    * R/newMethods.R: Revised assignments in plotting methods to be
    compatible with the Murrell/Sarkar preliminary trellis package.

    * src/nlmefit.c (internal_decomp): Corrected an error reported by
    Kathryn Chaloner where Srows >= ZXrows.

2001-01-10  Douglas Bates  <bates@stat.wisc.edu>

    * DESCRIPTION (Maintainer): Added Maintainer entry to
    DESCRIPTION.

2000-12-09  Douglas Bates  <bates@stat.wisc.edu>

    * R/lme.R: (Variogram.lme) Coerced n.pairs to numeric and
    corrected calls to dist.

2000-12-08  Douglas Bates  <bates@stat.wisc.edu>

    * R/lme.R: Changed all calls to dist in lme.R, gls.R and
    corStruct.R to use the argument "method" rather than "metric".

    * R/pdMat.R: Replaced calls to .Fortran("dbksl", ...) with calls
    to the R function solve() for R-1.2.0 where the S_compat.c file
    has been removed.

    * inst/scripts/ch01.R: Added the chapter scripts from
    "Mixed-Effects Models in S and S-PLUS" to the inst directory.

    * man/formula.nlme.Rd: Modified the examples in this and many
    other .Rd files to use the Loblolly data and the SSasymp
    self-starting function.  There were many occasions where the
    Soybean data was being fit to a default formulation of an nlme
    model just to get things like a formula.  This fit of the Soybean
    data is difficult and time consuming and not very illuminating.

2000-11-30  Douglas Bates  <bates@stat.wisc.edu>

    * src/matrix.c: Removed dependency on S_compat.h by replacing
    calls to dqrdca with the equivalent call to dqrdc2.

2000-07-24  Saikat DebRoy

    * Changed various uses of do.call to eval.  The reason for doing
    this is to allow calls to the function calling eval from within
    another function.  It's the old scoping problem again.

2000-07-07  Saikat DebRoy

    * src/nlme.c, src/nlmefit.c: Added mixed_grad in calls to optif9.

    * src/nlmefit.c : Added handling of pdLogChol objects (pdClass =
    4) in various functions.

    * src/nlmefit.c (mixed_grad): New function giving analytic
    gradient for mixed-effects log-likelihood.

    * R/nlme.R (nlme.formula): Added updating of the pdFactor in the
    PNLS step.

    * R/reStruct.R: Made pdLogChol the default representation for
    reStruct objects.

    * src/pdMat.c (logChol_pd): new function.

    * R/pdMat.R: Added back the pdLogChol class and associated
    methods.


2000-07-03  Douglas Bates  <bates@stat.wisc.edu>

    * man/corFactor.corStruct.Rd: Added an alias for
    corFactor.corNatural

    * src/nlmefit.c (internal_loglik): Changed the return value on
    error from R_NegInf to -DOUBLE_XMAX so the optif[09] optimizers
    will work properly.

    * src/corStruct.c (corStruct_factList): Provided a correct call to
    the R version of F77_CALL(chol) here and everywhere else in this
    source file.

    * R/zzz.R: Removed the deprecated call to provide

2000-03-17  Douglas Bates  <bates@stat.wisc.edu>

    * added many data sets

    * incorporated patches from Brian Ripley to use optim for the
    optimization rather than nlm

    * src/nlmefit.c (internal_loglik): Returned R_NegInf instead of
    generating a error on a singular precision matrix.

    * DESCRIPTION (Depends): Changed Depends to R version 1.0.1
    because of problem with assigning components of lists.

1999-12-23  Douglas Bates  <bates@stat.wisc.edu>

    * TITLE: Changed the title to nlme