annotateMosaic: Add tooltips to a Mosaic plot
In duncantl/SVGAnnotation: Tools for post-processing SVG plots created in R

Description Usage Arguments Value Note Author(s) References See Also Examples

The function annotateMosaic adds basic tooltips to a mosaic plot for visualizing frequency tables and conditional probabilities. The tooltips help viewers to comprehend what can be a complex plot. The tooltips currently give the count along with the different levels of the categorical variables. One can specify a function that is called to create the tooltip for each cell.

Two lower level functions are getCategoryLabelNodes.mosaic getAxesLabelNodes.mosaic. These return the SVG nodes corresponding to the labels for each category on the different sides of the plot and the SVG nodes for the axes/variable labels, respectively.

These functions annotate the SVG nodes with meta-informatio about the type. This allows us to retrieve the same nodes in subsequent calls even if we have modified the structure of the SVG document as we add markup such as tooltips and links.

Note that for the current implementation, if any "row" in the mosaic plot has all cells with zero observations, the subsequent tooltips will be mis-aligned, i.e. some of the latter tooltips will be for the wrong (earlier) cells.

annotateMosaic(doc, table, cellTips = NULL,
               axisLabelTips = c(names(attr(table, "col.vars")),
                                 names(attr(table, "row.vars"))),
               categoryTips = getMosaicLabels(table),
               shade = FALSE, addCSS = TRUE)
getCategoryLabelNodes.mosaic(doc, table = NULL,
                              numCells = prod(dim(table)), groupZeros = NA)
getAxesLabelNodes.mosaic(doc, table = NULL, groupZeros = NA)

`doc`	the SVG document, typically created via a call to `svgPlot` but this can also be the name of a file containing the SVG content or the already parsed document.
`table`	the frequency table returned by the call to `mosaic`. This contains the information about the counts and the labels for the variables and their categories/levels and the order in which they were plotted.
`axisLabelTips`	the text for the axes labels tooltips
`categoryTips`	tooltips for the labels on the axes that identify the different categories of each of the categorical variables.
`shade`	a logical value indicating whether the `shade` parameter was specified as `TRUE` in the call to `mosaic` which would then give rise to a legend. It is important to know this to interpret the nodes in the SVG content.

`addCSS`	a logical value indicating whether to add the (default) CSS to the SVG document or the name of a CSS file or CSS content to add.
`cellTips`	a character vector of tool tips for the cells in the plot or a function that is called to compute the tooltip string for a given cell. If this is a function it will be called with a single argument which is a data frame with one row giving the levels of the different categorical variables along with a variable named `.count` which gives the number of observations in that cell.
`numCells`	the number of cells in the mosaic plot. This can be specified instead of the `table`.
`groupZeros`	a logical value (or `NA`) indicating whether to perform the operation that groups SVG elements representing a cell with a zero count together into a single node. This makes processing the SVG easier.

The updated SVG document.

At present, we can only handle mosaic plots with less than 5 variables.

Duncan Temple Lang

The vcd package.

svgPlot

if(require(vcd)) {
  a = matrix(as.integer(rnorm(12, 100, 10)), 3, 4, dimnames = list(c("A", "B", "C"), c("W", "X", "Y", "Z")))
  names(dimnames(a)) = c("Var 1", "Var 2")
  saveXML(annotateMosaic(svgPlot(o <- mosaic(a)), o), "simpleMosaic.svg")



  saveXML(annotateMosaic(svgPlot( m <- mosaic(UCBAdmissions)), m), "ucb.svg")




  saveXML(annotateMosaic(svgPlot(z <- mosaic(Survived ~ ., data = Titanic)), z), "titanic.svg")




    # Example of 4 variables with more than 2 categories.
  xx = data.frame(a = sample(1:3, 100, TRUE), b = sample(c("A", "B", "C", "D"), 100, TRUE),
                  c = sample(c("X", "Y"), 100, TRUE), d = sample(c("Yes", "No"), 100, TRUE))

  saveXML(annotateMosaic(svgPlot(m <- mosaic(with(xx, table(a, b, c, d)))), m), "multiMosaic.svg")



    # Provide a function for creating the cell tooltip strings.
  saveXML(annotateMosaic(svgPlot(o <- mosaic(a)), o, cellTips = function(x) x$.count), "mySimpleMosaic.svg")



    # annotate the plot more manually, i.e. specifying links and
    # tooltips for the axes labels and category labels.

  data(UCBAdmissions)
  doc = svgPlot( m <- mosaic(UCBAdmissions))

  cats = getCategoryLabelNodes.mosaic(doc, m)
  ax <- getAxesLabelNodes.mosaic(doc, m)

#  annotateMosaic(doc, m, categoryTips = character())

  mapply(addLink, ax,
           c("http://www.math.yorku.ca/SCS/vcd/",
             "http://cran.r-project.org/web/packages/vcd/index.html",
             "http://www.omegahat.org/SVGAnnotation"))

  tips = c(Rejected = "The applicant was not admitted this year",
           Admitted = "The applicant was admitted ",
    F = "Department of ...",
    E = "Department of ...",
    D = "Department of ...",
    C = "Department of ...",
    B = "Department of ...",
    A = "Department of ...",
    Male = "Applicant was a man",
    Female = "Applicant was a woman")

  addToolTips(cats, tips[names(cats)])
  saveXML(doc, "iucb.svg")
} else {
  warning("You need the vcd package (Visualizing Categorical Data) for these mosaic examples")
}