version 1.2-33 2008-05-05

    glm.diag: 'gaussian' not 'Gaussian' for family name.
    cv.glm.Rd: typo

version 1.2-32 2008-03-30

    Add inst/CITATION file

version 1.2-31 2008-02-18

    print.simplex now returns its result invisibly.

version 1.2-30 2007-10-03

    po/R-ru.po: new file
    *.Rd: remove old \non_function{} markup.

version 1.2-29 2007-08-29

    R/*.q add licence/copyright statements
    melanona.Rd: correct the coding of 'sex'
    beaver.Rd, bootfuns.q cv.glm.Rd, EEF.profile.Rd: spelling corrections
    glm.diag.Rd: add  \link markup.

version 1.2-28 2007-06-12

    INDEX, man: change 'library' to 'package' and remove some references to
        S-PLUS.
    bootfuns.q: avoid abbreviating arguments, switch to seq.int, add
	drop=TRUE.
    DESCRIPTION: R >= 2.4.0 because of seq.int.

version 1.2-27 2006-11-29

    bootfuns.q : use control=NULL for deparsed calls.

version 1.2-26 2006-09-05

    smooth.f.Rd, tilt.boot.Rd: omit empty argument

version 1.2-25 2006-07-25

    censboot.Rd: unique() in mel.fun was not working correctly due to
    rounding error on some P4 Linux machines.

version 1.2-24 2005-12-09

    catsM.Rd: correct units for hearts to that given (correctly) in MASS.
    Add French and en@quot translation.

version 1.2-23 2005-07-26

    Add error message for a = NA in bca.ci.
    In cv.glm, eval updated formula in the parent.

version 1.2-22 2005-02-01

    Improve error messages for possible translation.

version 1.2-21 2005-01-21

    remove <<see below>> in ?boot.array

version 1.2-20 2004-11-03

    remove require(ts) call in tsboot

version 1.2-19 2004-08-12
version 1.2-18 2004-08-03

    versions for R 2.0.0.

version 1.2-17 2004-04-20

    remake datasets: looks like wool, coal and manaus were corrupt.

version 1.2-16 2003-12-08

    updates for R 1.9.0's namespaces

version 1.2-15 2003-10-22

    consistency in documentation of default args.
    tapply(foo)[bar] is now still an array in R.

version 1.2-14 2003-08-10

    remove unused vars
    add local definition for is.missing
    use \dQuote markup in help files.
    ensure .Random.seed is saved/retrieved in the workspace.
    remove/comment redundant assignments.

version 1.2-13 2003-03-09

    update NAMESPACE file
    censboot.Rd: call generic, not predict.smooth.spline

version 1.2-12 2003-03-01

    Use namespace, remove help for internal functions.
    glm.diag.plots had one set of axis labels reversed.

version 1.2-11 2003-01-29

    Replace inv.logit with overflow-proof plogis.
    Remove birthwt (from MASS) and help page for lynx.

version 1.2-10 2002-11-30

    Use optim(method="BFGS") for 1D optimization

version 1.2-9 2002-11-05

    Remove datasets taken from MASS package.

version 1.2-8 2002-04-11

    Remove arima.sim.
    Converted data sets to version 2 .rda, compressed larger ones.
    More care over .Random.seed.

version 1.2-7 2002-01-29

    Added some help pages from Angelo Canty.
    Bug fix to simplex() from Duncan Murdoch.

version 1.2-6 2001-11-27

    Start documenting rest of the objects and arguments.

version 1.2-5 2001-08-16
version 1.2-4 2001-08-08

    Use TRUE and FALSE on help pages, add ... to arg lists.

version 1.2-3 2001-06-16

    Change to survival from survival5.

version 1.2-2 2001/03/31

    Add priority: recommended.
    Remove test directory, as R CMD check works (but takes a long time).

version 1.2-1 2001/03/03

    Mods from Angelo Canty for version 1.2 of the library section.
    Dataset beaver was corrupted and it has been repaired.
    Change T, F to TRUE, FALSE in R code.

version 1.1-7 2000/12/28

    Changes for R 1.2.0, this is really boot version 1.1 now,

version 1.0-6 2000/11/27

    Re-port all the help files with Sd2Rd 1.15.
    R-specific change in glm.diag.


version 1.0-5 2000/02/06

    Change tsboot for R, make tsboot examples work.

    Revise censboot examples to avoid conflicts with survival5 datasets.

    Help files for datasets now use keyword `datasets' not `sysdata'

    Replace nlm by optim, as works much better. One example still
    fails, but it diverges on S and the authors did not notice.


version 1.0-4 1999/11/21
version 1.0-3

    Adjustments for later versions of R (removing functions added here).


version 1.0-2 1999/02/24

    Removed sample, model.response from zzz.R

    Changed attributes(out$spa)$name<- to names(out$spa)<- in saddle.

    Altered test/test-examples to make use of the examples now in R-ex.

    Note that saddle.distn and functions that depend on it (control)
    may fail due to lack of robustness of nlm.


version 1.0-1 1998/07/24

boot/R:

    Many funs store the .Random.seed but that need not exist, so call
    runif(1) first.

    Add missing function union in zzz.R

    Added version of sample that uses prob= in zzz.R

    Replace is.matrix by isMatrix <- function(x) length(dim(x)) == 2
    as R does not think a data frame is a matrix.

    Replace reading of a number in glm.diag.plots by readline.

    Replace nlmin by nlm, with appropriate changes to returned components.
    (This is not wholly successful, as nlm seems less tolerant.)

    Comment out assigns to frame=1.

    split.screen etc are not in R yet: use par(mfrow) in plot.glm.diag and
    layout in plot.boot.

    add drop=F in cv.glm, cens.weird.

    change is.inf to is.infinite.

    change maxit= in glm calls in saddle to calls to control=

boot/data:

    add a dummy rts:  rts <- function(units, name, ...) ts(...)
    to top of bootdata.q

    edit n= to nmax= in scan (carefully: also occurs in lists in scan
    commands).

    add

       for(obj in ls()) save(obj, file = paste(obj, ".rda", sep=""), ascii=T)

    to end of bootdata.q

       R -v 20 --no-save < bootdata.q

    cd4.nested is made by bd.q in the top directory.

boot/man:

    convert files by Sd2Rd. Add data() lines.

boot/test:

    script to test the examples in the help files. Not all will run,
    as far as I know due to things missing in R or nlm not coping
    as well as nlmin.