The US presidential elections has popularized “cartograms”, a type of map visualization that makes it easier to distinguish areas of higher numeric values (like higher vote counts) by exploding tiny polygons (like coastal areas in the East Coast) to better reflect their perceived values. For example, take a look at this map of the 2008 presidential elections:
Source: University of Michigan
Red states reflect higher vote counts for John McCain, and blue states reflect higher vote counts for Barack Obama. Now, when you “cartogram-ize” this map, you get the following result:
These cartograms have been around for quite a while, and are often used in GIS presentations to emphasize the point on how easy it is to create a false visual impression when authoring thematic maps on geographies that have a high contrast between the large and small polygons. This is especially true in US based maps, where coastal areas with high population counts are represented by tiny polygons.
I have often wondered how these cartograms are created, and was told by a graduate student at UCLA that there is now a java tool called ScapeToad that generates cartograms. Moreover, ScapeToad provides a free downloadable wizard-based tool that converts any shapefile into a cartogram! It is insanely cool. Here is a tutorial on how to use ScapeToad.
Step 1: Download ScapeToad
Go to http://scapetoad.choros.ch/download.php to download ScapeToad, and run the executable file.
Step 2: Add a layer
Run ScapeGoat, and click on the “Add layer” button, and choose a shapefile to upload. For this example, I chose a USA County shapefile that comes with basic 2000 census data.
Step 3: Create the cartogram
Now click on the “Create cartogram” button, and go through the wizard. For this example, I chose to create a cartogram based on the attribute for “Hispanic”.
Step 4: Export as Shapefile
Now that you have created a cartogram, you are ready to export it to something usable. ScapeToad allows you to export to SVG (which you can import into Illustrator), or as a shapefile (which you can import into ArcGIS). Click on “Export to Shape”, and save the cartogram as a shapefile. Open ArcGIS and load the shapefile.
When the shapefile is loaded into ArcGIS, it will be drawn in one random color, as this is the default behavior for loading any shapefile in ArcGIS. Right click on the cartogram layer, and modify the symbology. I chose to symbolize by:
- Quantities -> Graduated colors
- Value: Hispanic
- Classification: Natural Breaks
- Classes: 3
Now the cartogram is color coded based on the variable it was cartogramed with: Hispanic. As a final step, I chose to label the counties that have high values:
US counties with high hispanic population counts (Census 2000)
