A treemap displays hierarchical relationships through a set of rectangles, sized proportionately to each data point, clustered together into one large rectangle. The rectangular screen space is divided into regions, and then each region is divided again for each level in the hierarchy. Treemaps show part-to-whole relationships with each rectangle in the tree map representing a category from the dataset. The nested regions show hierarchical relationships and allow for quantitative comparisons of attribute values. A second variable for each category can also be coded using colour.
Proportionate aid given by groups in 1990
This treemap shows the proportionate aid given by each group in 1990. The slider across the bottom allows the user to change the year, which would reflect in changes to the proportions shown in the treemap.
Advice for CHOOSING this option
Because treemaps can quickly become unwieldy, consider choosing them when the size of each category is best represented by proportionate area. The smallest value should be no less than 5% of the size of the largest value, so that both are distinguishable.
Advice for USING this option
Depending on the range of quantities in the dataset, a treemap can become quite large. The smallest rectangles in the treemap may be too small to be labeled. In these instances, treemaps may be better suited for situations in which a viewer can interact with the display rather than view it in static print. In interactive situations, a viewer can hover over the smallest rectangles to reveal a popup box with the full details while in static situations the treemap could be uninterpretable. All values must be positive in order to use a treemap.
Treemaps: A Tool for Structuring, Exploring and Summarising Qualitative Information: This tool demonstrates how qualitative information can be displayed using a Tree Map.
Ten Lessons in Treemap Design: Excellent diagnosis and tips on creating treemaps from Juice Analytics.
Other ways to see the parts of a whole
Arranging a matrix of icons (usually 100 or 1000 icons) typically as a frequency-based representation, simultaneously displaying both the number of expected events and the number of expected non-events.
Illustrating proportion through a circular chart divided into sectors (like slices of a pie).