CHANGES in `multcomp' VERSION 1.1-2 o mcp may average over interactions and covariates (Experimental!) o ylim argument to plot o update to survival 2.35-7 CHANGES in `multcomp' VERSION 1.1-1 o better handling of model.frames in `cld' CHANGES in `multcomp' VERSION 1.1-0 o new compact letter displays o force binary mode when downloading SPSS data files CHANGES in `multcomp' VERSION 1.0-8 o new vignette containing supplementary material for a letter on Pollet \& Nettle (2009, Evolution and Human Behavior) o references updated CHANGES in `multcomp' VERSION 1.0-7 o new default for missing linfct: inference over all parameters as given by coef(model) o new function cftest, similar to lmtest:::coeftest CHANGES in `multcomp' VERSION 1.0-6 o better names for p-values (feature request by Achim Zeileis) CHANGES in `multcomp' VERSION 1.0-5 o update vignette CHANGES in `multcomp' VERSION 1.0-4 o fix Rd problems CHANGES in `multcomp' VERSION 1.0-3 o use isSymmetric(vcov, tol = sqrt(.Machine$double.eps) to check for symmetry of covariance matrices in `parm' (suggested by Fritz) o better error messages suggested by Rich CHANGES in `multcomp' VERSION 1.0-2 o use new model.matrix / model.frame methods for class mer in lme4 (>= 0.999375-16) CHANGES in `multcomp' VERSION 1.0-1 o make sure more than 2 groups are present in contrMat() o check if `model' is an S4 object before trying to extract its `call' slot o better warning when df.residual fails CHANGES in `multcomp' VERSION 1.0-0 o unify both vignettes o mcp() doesn't try to take care of interaction terms anymore CHANGES in `multcomp' VERSION 0.993-3 o new `parm' function for specifying model parameters estimated elsewhere o improve print method for confidence intervals o new adjustment method `free' (for free combinations) CHANGES in `multcomp' VERSION 0.993-2 o make sure names are printed correctly when nrow(K) = 1 o interaction terms are now correctly handled when a symbolic descripton of the hypotheses is supplied, thanks to Antonio Fabio Di Narzo for spotting. CHANGES in `multcomp' VERSION 0.993-1 o new vignette `Simultaneous Inference in General Parametric Models' o warn if ... arguments are ignored o levels in trees513 data slightly changed. CHANGES in `multcomp' VERSION 0.992-9 o remove deprecated functions simint and simtest (the files are still available from multcomp/inst/deprecated) o better error message when factors specified in mcp() aren't found in model CHANGES in `multcomp' VERSION 0.992-8 o add model.frame and model.matrix methods for class `lme' (feature request by Antonio Rausell ) o new data set `trees513' CHANGES in `multcomp' VERSION 0.992-7 o new contrasts `UmbrellaWilliams' and `GrandMean' (thanks to Daniel Gerhard ) CHANGES in `multcomp' VERSION 0.992-6 o add support for lme4::(g)lmer2 objects (thanks to Manuel Eugster) o coef. and vcov. didn't always pass through, spotted by John Deke CHANGES in `multcomp' VERSION 0.992-5 o remove non_function entries in Rd files CHANGES in `multcomp' VERSION 0.992-4 o update to mvtnorm 0.8-0 CHANGES in `multcomp' VERSION 0.992-3 o better support for `mer', `lmer' and `glmer' objects (package `lme4') CHANGES in `multcomp' VERSION 0.992-2 o add support for `lme' objects (feature request by John Wilkinson and Dieter Menne ) CHANGES in `multcomp' VERSION 0.992-1 o fix LaTeX problem in vignette (spotted by Brian D. Ripley) CHANGES in `multcomp' VERSION 0.991-9 o fix some minor problems spotted for plotting or printing of confidence intervals (thanks for Rich Heiberger and Ludwig Hothorn) CHANGES in `multcomp' VERSION 0.991-8 o make sure `type' is available o catch errors with `aovlist' objects o print for confint method indicates if confidence intervals are adjusted or univariate CHANGES in `multcomp' VERSION 0.991-7 o try to deal with non-estimable coefficients in `model': if the correspondings columns of `linfct' are zero anyway, we silently proceed. If not, we stop and ask for a correctly specified model. o try to detect interactions with numeric covariates. We need to choose a specific value here but it seems hard to add something appropriate to the user interface (I don't want to touch `mcp'). For the time being, we fire a warning and the user can modify `linfct' manually. o add mcp( = "Means"), contribution by Richard M. Heiberger o add Rich to DESCRIPTION file CHANGES in `multcomp' VERSION 0.991-5 o plot.confint allows for specification of xlim, xlab and main (patch by Rich Heiberger) CHANGES in `multcomp' VERSION 0.991-4 o Depends: stats o confint(..., calpha = x) can be used to specify critical values directly (with x being either a function or a scalar) CHANGES in `multcomp' VERSION 0.991-3 Implement feature requests by Rich Heiberger : o glht.mcp adds a `focus' element to the returned object containing the names of the factors under test o plot.glht calls plot.confint.glht o documentation updates o fixed three minor bugs CHANGES in `multcomp' VERSION 0.991-2 o mcp(fact = K) with K = contrMat(...) didn't work spotted by Yves Brostaux