Java Demos
  Rectangular Cartograms
Kinetic Collision Detection
Maximum Regression Depth

Rectangular Cartograms


This demo is a (very much) slimmed down version of a program that computes rectangular cartograms with the so-called segment moving heuristic (see the references below for details). This version supports only one map and one data set, namely Europe and its population. We allow six pairs of adjacent countries to have different relative positions. This leads to 64 different layouts, 32 of which are realizable. The country pairs are: France and Italy, Italy and Austria, Austria and Hungary, Hungary and Ukraine, Poland and Ukraine, and Serbia & Montenegro and Romania.

Instructions: Press Go after choosing your favorite set of parameters. The program will then try all realizable layouts and present the one with the smallest average error as output. Aspect Ratio bounds the maximal aspect ratio of the rectangles, Adjacencies decides if the program is allowed to destroy the correct adjacencies of the countries, and Show Error colors the countries according to the cartographic error (see the abstract below for details). If Animation is selected then the segment moving heuristic is animated from the best layout to the best cartogram.

Note: If Animation and Adjacencies: false is selected at the same time, then the program seems to pause somewhere in the middle of the animation. This is because it first tries to maintain the correct adjacencies until it can not make progress anymore. Only then is the heuristic allowd to push segments over adjacency boundaries.

A more extensive version of this demo is decribed in

Rectangular Cartograms: Construction & Animation
S. Florisson, M. van Kreveld, and B. Speckmann.
14th Annual Multimedia Review of Computational Geometry, Proc. 21st ACM Symposium on Computational Geometry, 2005 (to appear).

and many more (algorithmic) details can be found in

On Rectangular Cartograms
M. van Kreveld and B. Speckmann.
Proc. 12th European Symposium on Algorithms (ESA), pp. 724-735, Lecture Notes in Computer Science 3221, Springer Verlag, 2004.
This technical report contains a full version of the paper.

We extended this work with a linear programming based heuristic that allows us to compute larger cartograms. See

A Linear Programming Approach to Rectangular Cartograms
B. Speckmann, M. van Kreveld, and S. Florisson.
Proc. 12th International Symposium on Spatial Data Handling (SDH), 2006 (to appear).

for algorithmic details and for some cartograms.

Code by Sander Florisson, Marc van Kreveld, and Bettina Speckmann.

last modified: 02-May-2006