Version 0.9-0 initial release to CRAN

Version 1.0-0

o Rewrote summary.effect method and added print.summary.effect method.

Version 1.0-1

o Blanks can be inserted into or removed from effect names without causing an error; thus, e.g., "poly(education,3)" is equivalent to "poly(education, 3)".

o Name spaces of lattice and grid packages are imported, as required in R 1.8.0.

Version 1.0-2

o Added ask argument to plot.effect.list, and row, col, nrow, ncol, and more arguments to plot.effect, to support graphing an array of effect plots.

o Fixed bug in plot.effect that caused xlab argument to be ignored in certain circumstances.

Version 1.0-3

o effect function now works if na.action is na.exclude.

Version 1.0-4

o Fixed small bug introduced in version 1.0-3.

Version 1.0-5

o x.var and z.var arguments to plot.effect now take names as well as indices.

Version 1.0-6

o A variable specified in xlevels can be fixed to a single value.

Version 1.0-7

o Made effect() generic, with a method for lm objects that handles glm objects as well.

Version 1.0-8

o Small fixes to the help files.

Version 1.0-9

o Small change to compile without a warning in R 2.4.0.

Version 1.0-10

o Standard errors of effects are computed using t rather than standard-normal distribution for models with estimated dispersion (as suggested by Brian Ripley).

o Small fixes.

o Objects are now named "eff" and "eff.list" rather than "effect" and "effect.list".

o Data sets now provided by lazy data.

Version 1.0-11

o Replaced obsolete \non_function{} markup in Rd files (reported by Kurt Hornik).

Version 1.0-12

o key.args argument added to plot.eff() (coutesy of Michael Friendly), to allow conrol over, e.g., placement of legend.

Version 2.0-0

o Jangman Hong joins project.

o support added for multinomial and proportional-odds logit models, as fit by multinom() (in nnet package) and polr() (in MASS) package, following results in Fox and Andersen (2006).

o added the argument given.values to effect() methods for finer-grain control of displays.

Version 2.0-1

o Fixed bug in effect.polr() that prevented computation for a model with a single term (reported by Paul Prew).

Version 2.0-2

o Fixed bug in print(), summary(), and plot() methods for polytomous logit models with a response with numbered levels (reported by Paul Prew).

Version 2.0-3

o Fixed bug in all effect() methods that caused error when na.action="na.exclude" (reported by Tracy Lightcap and Rob Goedman).

Version 2.0-4

o Palettes from the colorspace package are used by default for stacked plots.

o Fixed bug in handling of typical= argument to effect() (argument was effectively ignored).

o Added Titanic and Wells data sets.

o Small changes.

Version 2.0-5

o Added examples for Titanic, BEPS, and WVS data sets.

o Arguments ... (e.g., digits) passed through in print() methods.

Version 2.0-6

o Fixed small bugs in print.efflist(), summary.efflist(), and plot.effpoly() methods.

o Corrected error in missing-data handling that sometimes surfaced in effect.multinom().

o Added .Rd file for package.

Version 2.0-7

o Fixed bug in handling of given.values argument to effect().

Version 2.0-8

o The S3 method print.summary.eff is now declared in NAMESPACE (as it should have been all along).

o Added CITATION.txt file (courtesy of Achim Zeileis).

o Version corresponding to John Fox, Jangman Hong (2009), Effect Displays in R for Multinomial and Proportional-Odds Logit Models: Extensions to the effects Package. Journal of Statistical Software, 32(1), 1-24 <http://www.jstatsoft.org/v32/i01/>.

O Fixed [pgk] markup in .Rd file cross-references.

Version 2.0-9

o Applied patches contributed by Ian Fellows to allow logical predictors and various coercions in model formulas to work properly.

o Fixed name of CITATION file (was CITATION.txt).

o Small changes to docs.

Version 2.0-10

o Backed out Ian Fellows's patches because of errors.

