Changes in version 0.9.6 - Fix @aliases for roxygen 8.0.0 - Make some color_table() examples visible in documentation Changes in version 0.9.5 This is a major release of the package, completing work on a sizable collection of tidy tools for manipulating categorical data in various forms - Added a general collapse_levels() function that can collapse levels of variables belonging to data sets of any form. [GK] - Added as_matrix() to the set of as_*() conversion functions. [GK] - Added prop arguments to applicable as_*() conversion functions to easily convert counts to proportions (either relative to the grand total count or to specified margins). [GK] - Added vignette on tidy operations (convert, collapsing) [GK] - Gavin Klorfine (@gklorfine) becomes a package author Changes in version 0.9.4 - suppressWarnings() from ca:mcja() in mcaplot() examples Changes in version 0.9.3 (2026-03-18) This bundle brings quite a few enhancements, improvements in documentation and bug fixes - Added tidy conversion functions: as_array(), as_caseform(), as_freqform, as_table() PR #22 [Thx: Gavin Klorfine] - Fixed bug in mcaplot() coming from `ca::cacoords(): "non-conformable arguments" - Expanded documentation of color_table() to give better advice on how to use this in Rmd or qmd documents. - Added: knitr_include() as a general solution to using gt, DT, plotly, ... outputs in non-HTML documents. - Added pairs_diagonal_mosaic(), overriding the {vcd} version to give more flexibility in printing the cell values in the diagonal cells. PR #24 [Thx: Gavin Klorfine] Changes in version 0.9.2 - Added a label = c("name", "formula") argument to LRstats() to provide for labeling models by their model formulas in the output using get_models(). - Handle list (...) of models with formula labels more flexibly in LRstats() - Document get_model() and get_models() together - Added `assoc_graph() and a plot method for association graphs of loglinear models. - Added edge weights to assoc_graph() representing partial G^2 or Cramer's V Changes in version 0.9.1 (2026-02-09) This is a major release of the package, adding facility to show observations in mosaic plots as jittered points and a new visualization method based on background shading of frequency tables. - Added labeling_points() for mosaic displays to show observed or expected frequencies as random points in the tiles - Changed defaults for clip and margin in labeling_points() to be more sensible [Thx: Achim Zeileis] - Added color_table() to display frequency tables with color shaded backgrounds to show patterns - Refactored color_table() as S3 methods, adding support for frequency data frames. Added file output, other formatting options - Print message for color_table() giving the model fit statistics. - Now provide color_table(values =) to display either the frequencies (default) or residuals in the table. - Add legend = "note" to include a table note regarding what is shown. - Added Reinis data as an example of a higher-way table, 2^6 - Make the default label for seq_loglm() models reflect the model type - Added get_models() to extract model formulas from loglmlist and glmlist objects - Added get_model() to do the same for loglm() and glm() objects - Fixed bug in get_models() where the abbrev argument caused an error Changes in version 0.9.0 (2026-01-21) This is a major release of the package, adding structure to dataset examples and introducing an new extension Woolf's test for homogeneity of odds ratios. - Begin to categorize the datasets in the package with @concept tags - Begin to flesh out more general conception of CDA - datasets(): added ndim arg; package="vcdExtra" default - Added a tidyCat hex logo - Generalized vcd::woolf_test() to handle strata better, with special handling for 2 x 2 x R x C case. - Completed extension of vcd::woolf_test() to fully handle R x C strata, with tests for rows/cols/residuals - Added GH actions R-CMD-check & pkgdown - Now depends R (>= 4.1.0) due to use of native pipe |> - Fixed some broken URLs in R/data.R Changes in version 0.8.7 (2025-12-11) This is a major release of the package, fixing bugs and revising documentation o Added tests for CMHtest() PR #13 [Thx: Daniel Sabanes Bove] o Automatically omit strata with a single observation in CMHtest() because they do not contribute to the test statistics o Use the generalized Moore-Penrose inverse from MASS in CMHtest() such that it can work when the variance matrix is singular. o Converted the package to use roxygen documentation via {rd2roxygen}. UGH! o Added CrabSatellites data from {countreg} b/c that's still not available on CRAN o Moved grid from Depends: to Imports: o Now use markdown in package documentation for easier maintenance, via {roxygen2md} Changes in version 0.8-6 (2025-07-24) This is a minor release to satisfy the CRAN gods who like all crossref links to be correct. o Fixed CRAN nits re crossrefs o Update issues link o Update README o Fixed Winbuilder notes re: Rd links to data sources Changes in version 0.8-5 (2023-08-22) - Fix CRAN nit re vcdExtra-package.Rd - Fix moved URL - http: -> https: - Fix xrefs in vignettes - More spellcheck - Fixed one more NOTE re: AirCrash.Rd for a URL Changes in version 0.8-4 (2023-04-16) - Imports: tidyr must depend on (>= 1.3.0) for one use in a vignette. - Enhanced the vignette, "Creating and manipulating frequency tables", vignettes/creating.Rmd - Created a new vignette, "Mobility tables", with extensive examples of models and graphs for square mobility tables. Changes in version 0.8-3 (2023-03-04) - All datasets now classified with \concept{} tags by method of analysis. These can be found using help.search(pattern, field="concept") - Added a vignette, datasets.Rmd showing all datasets classified by method tags. Links only work in the pkgdown site. - Added Asbestos data Changes in version 0.8-2 (2023-01-29) This is a major enhancement release of the vcdExtra package, focusing on documentation and examples. - added HouseTasks data set, illustrating permutation of row / col variables - package now depends on R (>= 3.5.0) per CRAN nit - add Suggests: seriation to illustrate CA re-ordering of rows/cols based on correspondence analysis - all .Rd files reformatted and many examples extended. - the vignette mosaic.Rmd on mosaic displays has been extensively revised with examples for square tables and permutation of row / column variables. - begin to classify datasets with \concept{} tags - Added a new demo-housing.Rmd vignette, using content from demo/housing.R. Changes in version 0.8-1 - rename vignettes to be in order Changes in version 0.8-0 (2022-04-21) - Fixed warning from expand.dft() re type.convert - Old .Rnw vignettes converted to .Rmd - Fixed two problems detected in the initial submission. Changes in version 0.7-6 (2022-02-14) - Fix some issues with CMHtest() types argument #PR11 [Thx: ShuguangSun, Matt Kumar] - Fix some Winbuilder URL nits; extensive spell checking Changes in version 0.7-5 (2021-01-25) - Fix problem re use of rgl in mosaic3d() examples Changes in version 0.7-4 - Fix datasets() to work with packages not using LazyData #PR7 [Thx: Duncan Murdoch] - Bump package Version Changes in version 0.7-3 - fix Version number for gnm in DESCRIPTION - fix expand.dft and expand.table to work with tibbles [Thx: Duncan Murdoch] - vcdExtra gets a hex sticker Changes in version 0.7-1 (2017-09-29) - Fixed buglet in expand.dft() when table is 1-dim [Thx: Long Qu] - Added zero.test(), a simple score test for zero inflation - Development has moved to https://github.com/friendly/vcdExtra Changes in version 0.7-0 (2016-02-25) - Added Glass data - introduce links and references to DDAR - added mcaplot() to plot MCA solutions in DDAR style - added update.xtabs() method - updated vignette to refer to DDAR Changes in version 0.6-12 (2015-12-23) - Added Burt data - Fixed examples/vignette for ggplot_2.0.0 Changes in version 0.6-11 (2015-09-17) - bump pkg Version for CRAN Changes in version 0.6-10 - Added HospVisits data Changes in version 0.6-9 (2015-07-11) - Added Mice data - Removed uses of summarise() and Summarise() from demos and examples, in preparation for deprecating them. - summarise() is now deprecated - Now use importsFrom() for all functions from recommended packages Changes in version 0.6-8 (2015-04-16) - Fixed Title: and Description: for CRAN Changes in version 0.6-7 - Removed loddsratio (now in vcd) (rev 252) - Removed print.Kappa (now in vcd) (rev 253) - Fixed bug in CMHtest(): rmeans and cmeans labels were reversed (rev 254) - Fixed error in Fungicide.Rd, now that we require vcd_1.3-3 (rev 254) - Added WorkerSat data - 2 x 2 x 2 - Added AirCrash data Changes in version 0.6-6 - Minor doc changes Changes in version 0.6-5 (2015-01-03) - Added Cormorants data (fixed to latest) - Added LRstats(), to replace Summarise. Older summarise() and Summarise() will eventually be deprecated. - Now Suggests: AER for NMES1988 data - collapse.table() now works with array objects Changes in version 0.6-3 (2014-11-01) - Fixed bug in logLik.loglm when the data contain zero frequencies (rev 228) - Made Summarise generic, adding a method for "glmlist" objects (rev 230) - Added a coef() method for "glmlist" objects (thx: John Fox) - Added a Summarise.loglmlist method (rev 232) - Replaced all documentation uses of summarise() with Summarise() (rev 233) - Added cutfac(), a convenience wrapper for cut() (rev 234) - Now use rgl:: in all mosaic3d functions Changes in version 0.6-2 - added Summarise, to replace summarise - Added HairEyePlace data - 4 x 5 x 2, Caithness and Aberdeen hair/eye color (rev 223) - Added PhdPubs data from Long (1997) - publications by PhD candidates - Allow Summarise to work with models w/o a deviance() function - Fixed bug in Summarise wrt degrees of freedom Changes in version 0.6-1 (2014-06-10) - Added ICU data - Added Toxaemia data - multivariate response contingency table (rev 209) - Added Vietnam data - 2 x 5 x 4 frequency table (rev 210) - Added logLik.loglm to allow use of AIC() and BIC() for loglm models (rev 212) - Fixed loddsratio.Rd to work with revised vcd::CoalMiners data (rev 212) - Added blogits for bivariate binary response data - Added Vote1980 data (rev 214) Changes in version 0.6-0 (2014-03-08) - Removed Authors: in DESCRIPTION, bumped Version Changes in version 0.5-12 - Added ShakeWords data set- word frequency counts from Shakespeare (rev 188) - Added Geissler data-- all family sizes for Saxony sex composition data (rev 190) - Added logseries functions for the logarithmic series distribution (rev 191) - Added Depends data -- dependencies of r packages (rev 192) - Fixed buglet in seq_loglm() not respecting arrays - Added seq_mosaic() (rev 194) - Added CyclingDeaths data (rev 196) - mosaic3d() gets an interpolate= option to control shading levels (rev 197) - Fixed bug in seq_mosaic, thx to David Meyer (rev 200) - Fixed bug in seq_loglm() when marginals != 1:nf Changes in version 0.5-11 (2013-09-16) - Added mosaic.glmlist to plot mosaics (or other strucplots) for some or all models in a glmlist (rev 169) - Added loglin-utilities.R, containing a suite of functions to provide a more conceptual way to specify loglinear models by type ('joint', 'conditional', 'mutual', 'markov', 'saturated') (rev 171) - Added mosaic.loglmlist, similar to mosaic.glmlist for models fit using MASS::loglm (rev 173) - Both mosaic.glmlist and mosaic.glmlist get an explicit panel=argument; both get some more sensible default default arguments (rev 175) - Added seq_loglm to fit sequential loglm models to marginal subtables, giving a loglmlist result (rev 176) - Added Accident data (rev 178); fleshed out Accident examples (rev 180) - Fixed use of ::: for R 3.0.1 (rev 179) - Fixed various problems related to use of MASS:loglm (rev 181-183) - Added Titanicp to datasets (rev 185) Changes in version 0.5-8 - Revised vcd-tutorial showing some examples of plyr; added a section on RC models - Added Donner data with example of ggplot2 plot for a binomial glm() - Added vcd-tutorial section using ggplot2 for Donner data - Enhanced datasets() to provide a maxTitle argument (rev 153) - Added doubledecker plots to Dyke.Rd (rev 156) - Added Draft1970table and Draft1970 data sets (rev 158) - Added example of doubledecker plots to vcd-tutorial vignette (rev 164) Changes in version 0.5-7 (2013-03-05) - Completed CMHtest methods, adding overall tests across strata in a general way - CMHtest now gets an S3 generic with a formula interface - print Kappa gets digits= and CI= arguments Changes in version 0.5-6 (2013-02-28) - Added Hosmer Lemeshow and HLtest methods, including plotting via vcd::rootogram() - Added CMHtest for general Cochran-Mantel-Haenszel tests - Revised vcd-tutorial vignette, adding a section on CMH tests; removed dependence on Z.cls Changes in version 0.5-3 (2012-05-08) - Added Mammograms data (4x4, ordered factors, agreement) - Extended mosaic.glm examples - Added Alligator data (4x2x2x5, in frequency form) - Added DaytonSurvey data (5-way, 2x2x2x2x2 in frequency form) - Extended vcd-tutorial vignette with a section on collapsing over factors - Removed aperm.* now that aperm.table is in base R Changes in version 0.5-2 (2011-03-24) - Added loddsratio and related methods for log odds ratios, generalizing vcd::oddsratio from 2 x 2 (x strata) tables to R x C (x strata) tables - Added as.matrix.loddsratio, as.array.loddsratio methods - Added some simple plot examples to example(loddsratio), anticipating a plot method - Added data(Fungicide), a 2 x 2 x 2 x 2 table - Renamed summarize() and related methods to summarise() to avoid conflict with plyr. - Addition to vcd-tutorial vignette on use of aperm() with table objects - Updated demo(yamaguchi-xie) to correct row/col nomenclature and add plot of BIC - Added aperm() S3 generic to handle table objects - Moved tv.dat to inst/doc/extdata to avoid warnings in R 2.12+ Changes in version 0.5-1 (2010-11-29) - Added Yamaguchi87 data (5x5x3 three-way mobility table in frequency form) - Added demo(yamaguchi-xie) illustrating fitting and visualization of the models of homogeneous and log multiplicative layer effects fit in Xie (1992, Table 1) - Added BIC to summarize() and friends - Added Hauser79 data (two-way mobility table), plus some examples from Powers and Xie (2008) - Added Crossings() to construct interactions for Goodman 1972 crossings model. - Added datasets() to list datasets in packages - Extended description and examples of Kway() - Added meanResiduals() and extended mosaic.glm() (Heather Turner) - summarize() gets a sortby argument for glmlist and loglmlist objects Changes in version 0.5-0 (2010-04-16) - Fleshed out mosaic3d, allowing display of observed or expected, internally calculated or externally supplied residuals, specifying the initial 3D shape, etc. This completes the 'top-level' work on mosaic3d(), borrowing code from vcd::strucplot. - Added initial handling for zero cells in the table to mosaic3d(). - Added center3d() for finding the mean coordinates of shape3d objects. - Added demo(mosaic-hec) comparing 2D and 3D mosaics for HairEyeColor data - Gave mosaic3d a label_edge argument, allowing labels for dimensions at minima or maxima - Made mosaic3d object oriented, giving it a loglm method - Added Kway(), fitting all 0-way, 1-way, 2-way, ... k-way models in a glm Changes in version 0.4-3 - Added demo(mosaic3d-demo), a proof-of-concept for doing 3D mosaic displays - Added mosaic3d(), an initial basic Version. - Factored out split3d() and gave it S3 methods Changes in version 0.4-2 (2010-03-18) - Revised vignette("vcd-tutorial"): added some hints for mosaic(), corrected stuff regarding prior limitations of mosaic.glm() - Added demo(Wong3-1): three-way table, with models of conditional association - Added Suggest: effects for effects plots of glm(), multinom() and polr() models - Added demo(housing): visualize models fit in example(housing, package="MASS") using mosaic() and effect plots. - Updated demo(Wong2-3): added model comparison plots, glmlist processing - Added Suggest: VGAM - Extended package description in vcdExtra-package.Rd - Added glmlist() to facilitate processing, extraction, plotting, etc. of a collection of glm() models - Added loglmlist(), for collections of loglm() objects - Added summarize methods for glm, glmlist, loglm and loglmlist objects Changes in version 0.4-1 (2010-03-07) - Added example(Caesar), illustrating structural zeros - Re-named Heckman variables to e1971, ..., e1968 (errors from loglm); began example(Heckman) - Added example(Detergent), example(Dyke) - Fixed bug with mosaic.glm when data in global environment (Heather Turner) - Added sieve.glm and assoc.glm methods (MF) - Added modFit.glm and modFit.loglm - Added demo(Wong2-3) Changes in version 0.4-0 (2010-02-28) - Added new datasets: data/{Abortion, Bartlett, Caesar, Cancer, Detergent, Dyke, Gilby, Heart, Heckman, Hoyt, Mobility} from mosdata.sas via md2r.sas converter. - Fixed small documentation warnings - Switched inst/CHANGES to NEWS Changes in version 0.3-6 - Added Depends: gnm - Added demo/{mental-glm, ucb-glm, vision-quasi} - Added demo/{occStatus,yaish-unidiff} - Initial release to CRAN Changes in version 0.3-5 - mosaic.glm now uses object$data if available Changes in version 0.3-4 - Fixed bugs in mosaic.glm, mosaic.gnm in models with terms like Diag(dest, origin) that get included in x$xlevels Changes in version 0.3-3 - Fixed bugs in mosaic.gnm - Fixed print.GKgamma - Added example of GKgamma to vcd-tutorial Changes in version 0.3-2 - Added more examples to mosaic.glm.Rd Changes in version 0.3-0 - Fixed bugs in mosaic.glm and mosaic.gnm Changes in version 0.2 - Added vcd-tutorial vignette Changes in version 0.1 - Initial Version on R-Forge.