vcd (version 1.4-12)

mosaic: Extended Mosaic Plots


Plots (extended) mosaic displays.


# S3 method for default
mosaic(x, condvars = NULL,
  split_vertical = NULL, direction = NULL, spacing = NULL,
  spacing_args = list(), gp = NULL, expected = NULL, shade = NULL,
  highlighting = NULL, highlighting_fill = rev(gray.colors(tail(dim(x), 1))),
  highlighting_direction = NULL,
  zero_size = 0.5, zero_split = FALSE, zero_shade = NULL,
  zero_gp = gpar(col = 0), panel = NULL, main = NULL, sub = NULL, ...)
# S3 method for formula
mosaic(formula, data, highlighting = NULL,
  ..., main = NULL, sub = NULL, subset = NULL, na.action = NULL)


The "structable" visualized is returned invisibly.



a contingency table in array form, with optional category labels specified in the dimnames(x) attribute, or an object of class "structable".


vector of integers or character strings indicating conditioning variables, if any. The table will be permuted to order them first.


a formula specifying the variables used to create a contingency table from data. For convenience, conditioning formulas can be specified; the conditioning variables will then be used first for splitting. If any, a specified response variable will be highlighted in the cells.


either a data frame, or an object of class "table" or "ftable".


an optional vector specifying a subset of observations to be used.


a function which indicates what should happen when the data contain NAs. Ignored if data is a contingency table.


size of the bullets used for zero entries (if 0, no bullets are drawn).


logical controlling whether zero cells should be further split. If FALSE and zero_shade is FALSE, only one bullet is drawn (centered) for unsplit zero cells. If FALSE and zero_shade is TRUE, a bullet for each zero cell is drawn to allow, e.g., residual-based shadings to be effective also for zero cells.


logical controlling whether zero bullets should be shaded. The default is TRUE if shade is TRUE or expected is not null or gp is not null, and FALSE otherwise.


object of class "gpar" used for zero bullets in case they are not shaded.


vector of logicals of length \(k\), where \(k\) is the number of margins of x (default: FALSE). Values are recycled as needed. A TRUE component indicates that the tile(s) of the corresponding dimension should be split vertically, FALSE means horizontal splits. Ignored if direction is not NULL.


character vector of length \(k\), where \(k\) is the number of margins of x (values are recycled as needed). For each component, a value of "h" indicates that the tile(s) of the corresponding dimension should be split horizontally, whereas "v" indicates vertical split(s).


spacing object, spacing function, or corresponding generating function (see strucplot for more information). The default is spacing_equal if x has two dimensions, spacing_increase for more dimensions, and spacing_conditional if conditioning variables are specified using condvars or the formula interface.


list of arguments for the generating function, if specified (see strucplot for more information).


object of class "gpar", shading function or a corresponding generating function (see details and shadings). Components of "gpar" objects are recycled as needed along the last splitting dimension. Ignored if shade = FALSE.


logical specifying whether gp should be used or not (see gp). If TRUE and expected is unspecified, a default model is fitted: if condvars (see strucplot) is specified, a corresponding conditional independence model, and else the total independence model.


optionally, an array of expected values of the same dimension as x, or alternatively the corresponding independence model specification as used by loglin or loglm (see strucplot).


character vector or integer specifying a variable to be highlighted in the cells.


color vector or palette function used for a highlighted variable, if any.


Either "left", "right", "top", or "bottom" specifying the direction of highlighting in the cells.


Optional function with arguments: residuals, observed, expected, index, gp, and name called by the struc_mosaic workhorse for each tile that is drawn in the mosaic. index is an integer vector with the tile's coordinates in the contingency table, gp a gpar object for the tile, and name a label to be assigned to the drawn grid object.

main, sub

either a logical, or a character string used for plotting the main (sub) title. If logical and TRUE, the name of the data object is used.


Other arguments passed to strucplot


David Meyer


Mosaic displays have been suggested in the statistical literature by Hartigan and Kleiner (1984) and have been extended by Friendly (1994). mosaicplot is a base graphics implementation and mosaic is a much more flexible and extensible grid implementation.

mosaic is a generic function which currently has a default method and a formula interface. Both are high-level interfaces to the strucplot function, and produce (extended) mosaic displays. Most of the functionality is described there, such as specification of the independence model, labeling, legend, spacing, shading, and other graphical parameters.

A mosaic plot is an area proportional visualization of a (possibly higher-dimensional) table of expected frequencies. It is composed of tiles (corresponding to the cells) created by recursive vertical and horizontal splits of a square. The area of each tile is proportional to the corresponding cell entry, given the dimensions of previous splits.

An extended mosaic plot, in addition, visualizes the fit of a particular log-linear model. Typically, this is done by residual-based shadings where color and/or outline of the tiles visualize sign, size and possibly significance of the corresponding residual.

The layout is very flexible: the specification of shading, labeling, spacing, and legend is modularized (see strucplot for details).

In contrast to the mosaicplot function in graphics, the splits start with the horizontal direction by default to match the printed output of structable.


Hartigan, J.A., and Kleiner, B. (1984), A mosaic of television ratings. The American Statistician, 38, 32--35.

Emerson, J. W. (1998), Mosaic displays in S-PLUS: A general implementation and a case study. Statistical Computing and Graphics Newsletter (ASA), 9, 1, 17--23.

Friendly, M. (1994), Mosaic displays for multi-way contingency tables. Journal of the American Statistical Association, 89, 190--200.

Meyer, D., Zeileis, A., and Hornik, K. (2006), The strucplot framework: Visualizing multi-way contingency tables with vcd. Journal of Statistical Software, 17(3), 1-48. tools:::Rd_expr_doi("10.18637/jss.v017.i03") and available as vignette("strucplot", package = "vcd").

The home page of Michael Friendly ( provides information on various aspects of graphical methods for analyzing categorical data, including mosaic plots. In particular, there are many materials for his course “Visualizing Categorical Data with SAS and R” at

See Also

assoc, strucplot, mosaicplot, structable, doubledecker


Run this code

## Formula interface for tabulated data plus shading and legend:
mosaic(~ Sex + Age + Survived, data = Titanic,
  main = "Survival on the Titanic", shade = TRUE, legend = TRUE)

mosaic(HairEyeColor, shade = TRUE)
## Independence model of hair and eye color and sex.  Indicates that
## there are significantly more blue eyed blond females than expected
## in the case of independence (and too few brown eyed blond females).

mosaic(HairEyeColor, shade = TRUE, expected = list(c(1,2), 3))
## Model of joint independence of sex from hair and eye color.  Males
## are underrepresented among people with brown hair and eyes, and are
## overrepresented among people with brown hair and blue eyes, but not
## "significantly".

## Formula interface for raw data: visualize crosstabulation of numbers
## of gears and carburettors in Motor Trend car data.
mosaic(~ gear + carb, data = mtcars, shade = TRUE)

mosaic(PreSex, condvars = c(1,4))
mosaic(~ ExtramaritalSex + PremaritalSex | MaritalStatus + Gender,
       data = PreSex)

## Highlighting:
mosaic(Survived ~ ., data = Titanic)

mosaic(Improved ~ Treatment | Sex, data = Arthritis, zero_size = 0)
mosaic(Improved ~ Treatment | Sex, data = Arthritis, zero_size = 0,
       highlighting_direction = "right")

Run the code above in your browser using DataLab