A mosaic plot is used
to display relative frequencies for categorical variables. Each crossing
of the categorical values is represented by a tile. The area of each
tile is proportional to the frequency of that crossing. The graph
is the result of an iterative process. The first iteration splits
the plot area into tiles along the width according to the relative
frequency of the first category column values. Subsequent iterations
split the tiles from the previous iteration in the direction orthogonal
to the previous split using the relative frequencies of each category
column's values. By default, the gap (or gutter) for each split gets
progressively smaller down to a minimum gap of 3 pixels.
You
can use the GUTTER= option to specify a different gap for each split.
The following figure
provides an example for a three-way mosaic plot, which has three categories.
Note: The plot contains equal-sized
tiles for demonstration purposes.
In the example plot,
the first split is along the width for CATEGORY 1. The second split
is along the height for CATEGORY 2. Finally, the third split is along
the width for CATEGORY 3. Notice how the gaps between the tiles get
progressively smaller from the first split to the last split.
Note: The MOSAICPLOTPARM statement
can be use in GRIDDED, LATTICE, and REGION layouts only.