https://github.com/cran/HH
Raw File
Tip revision: 2ff3462ba5f3ddb02d664643554c6988dd56fe85 authored by Richard M. Heiberger on 13 September 2014, 08:34:04 UTC
version 3.1-5
Tip revision: 2ff3462
NEWS
Version 3.1-4
-------------

* 2014 September 11

* New function interval.glm for confidence and prediction intervals for glm objects.

* add odds and antiodds functions

* new useOuterStripsT2L1 function, based on latticeExtra::useOuterStrips

* \tabular{}{} in ancovaplot.Rd

* add between= argument to ancovaplot

* change mmc.Rd and other mmc help files to use mmcplot in the
  examples instead of the older HH:::plot.mmc.multicomp

* HHscriptnames now takes an edition argument

* Sent to CRAN


Version 3.1-3
-------------

* 2014 July 18

* move First Edition scripts from scripts/xxxx.R to scripts/hh1/xxxx.r
  in parallel with scripts/hh2/xxxx.R for Second Edition scripts.

* fix naming bug of reshape2::melt arguments in likert.formula

* Limited release to colleague

Version 3.1-2
-------------

* 2014 July 15

* repair sequencing of ar:p,ma:q for two-digit p or q in tsdiagplot.

* Limited release to colleague


Version 3.1-1
-------------

* 2014 July 09

* mmc.explain moved to an unexported function

* Include mmcplot and related functions (lattice replacement
  for plot.mmc.multicomp and plot.multicomp).

* unrecommend plot.mmc.multicomp() and plot.multicomp() in favor of
  mmcplot().  I am not deprecating them now.

* new hov and OneWayVarPlot

* new ancovaplot and related functions.

* unrecommend ancova() in favor of aov() and ancovaplot().  I am not deprecating it now.

* replace demo("ancova") based on the ancova() with newer version based on ancovaplot().

* add new demo("appleAncova") with a response, covariate, treatment factor, and blocking factor.

* merge matrixTrellis functions

* merge regressionDiagnostics functions

* merge dicreteUniform functions

* copy HH second edition files to scripts/hh2

* Limited release to colleague


Version 3.1-0
-------------

* 2014 July 01

* Start merging small packages written for the HH second edition
  into the HH package

* Include noop versions of hhpdf hhdev.off hhcapture hhcode
  (needed for HH2 script files)

* Modify mmc.default to include result$none$glht$focus and
  result$lmat$glht$focus

* freeze on 2014-07-03


Version 3.0-5
-------------

* 2014 June 28

* Adjust ylim in demo(ancova)

* mmc (multicomp.reverse) warning() changed to cat().
  This is informative and does not require user action.
  It should appear at the place where the function is called.

* data(R282.y) now used in HH2 instead of data(R282).  For the second
  edition of HH, the dummy variables previously stored in R282 are now
  generated in the code.

* demo(betaWeightedAverage) added.

* added  RcmdrPlugin.HH to Suggests:

* corrected abline in plot.summaryHH.regsubsets

* repair as.multicomp.glht to force height to be a vector.

* repair mmc.default to allow estimate.sign to be honored by lmat,
  and order.contrasts to be honored by lmat and none.

* give as.multicomp.glht default values for focus and height arguments

* ancova "main" argument handles label correctly now.

* new argument "name.prefix" to functions ladder.f and ladder.fstar.

* repair abind call in plotOddsRatio.base for R.
  R requires the a= argument, S-Plus acted as if the default was a=0.

* new plotOddsRatio function based on lattice.  New function returns a lattice object.
  Old plotOddsRatio.base function returns OddsRatio(x, alpha[1]).

* resave crime.rda with names(dimnames())

* add jury.rda and oral.rda to data directory

* new example to objip: {search(); objip()[2:3]}

* freeze on 2014-07-01


Version 3.0-4
-------------

* 2014 March 10

* in likert, arguments xlimEqualLeftRight xTickLabelsPositive are now ignored
  when both at and labels are explicitly specified.

* the likert paper has been published in the Journal of Statistical Software.  See
  ?likert for the reference.

* published on CRAN 2014-03-11


Version 3.0-3
-------------

* 2014 February 5

* correct typo in plot.likert.formula

* document formula restriction in man/likert.Rd

* document placing x-axis labels on the top axis for horizontal likert plots

* make changes to allow y-axis labels on the right axis for vertical
  likert plots and document it.

* restore old par.settings at the completion of
  panel.bwplot.intermediate.hh, remove S-Plus features.

* new argument width.cutoff in print.mmc.multicomp

* AElogrelrisk previously depended on the dataset being sorted
  order(PREF, RAND), but it wasn't documented.  logrelrisk had the
  same dependence and it was documented, but not enforced by the
  function.  Now both AElogrelrisk and logrelrisk do the ordering.
  The user no longer needs to worry about sort order.

* revised three demo files for better appearance of the plots.

* new function panel.bwplot.superpose

* published on CRAN 16-Feb-2014


Version 3.0-2
-------------

* 2014 January 6

* add position() to factors in cc176 dataset

* remove data files in text form and in S-Plus form

* fix typos in plot.multicomp.Rd mmc.Rd

Still needed
-------------

* Still needed: stop requiring dimnames in as.MatrixList.array (?).

* MMC rewritten in lattice: mmcplot, mmcmatch, mmcboth.
  Currently in separate mmcplot package

* HH2writing: hhpdf, hhdev.off, hhcapture need to be moved to HH as no-ops.
  Stangle.HH not needed in HH.
  Currently in separate HH2writing package

* new panel.bwplot.intermediate

* verify all scripts from First Edition


Version 3.0-1
-------------

* 2014 January 6

* direct copy of the unreleased HH_2.3-43

* The HH package is now at version 3.x-y to support the Second Edition
  of the HH book, Statistical Analysis and Data Display.  All examples
  from the CRAN HH_2.3-42, and scripts supporting the First Edition, continue
  to work.  New features are added.  The package is now supported only for R.
  Not all S-Plus specific contructions have been removed.

* ladder: no longer uses panel.cartesian.R and grid.axis.hh.R

* likert.formula: changes in rightAxis to rowSums are now correct in each panel.

* Fix up lots of .Rd files where the lines were too long, but not
  detected by R CMD check.

* Switch IMPORTS from reshape package to reshape2 package.
back to top