Below each map or cartogram displayed on this website there is a download button. By clicking it, you can download an SVG image containing the map. SVG files contain vector images, and can be edited and converted by Adobe Illustrator or Inkscape, a free, open source alternative. If you would simply like to convert your SVG map to a PNG image to use in a document or paper and don’t have access to these programs, you can use this website.
Yes. Images generated by go-cart.io can be distributed under a permissive liccense (CC-BY, see Creative Commons license): "This license lets others distribute, remix, adapt, and build upon your work, even commercially, as long as they credit you for the original creation." Please credit our work by citing: Gastner MT, Seguy V, More P. Fast flow-based algorithm for creating density-equalizing map projections. Proc Natl Acad Sci USA 115(10):E2156–E2164 (2018). If possible, please also include a hyperlink to go-cart.io.
Cartograms are designed to visualize totals, not indices. In general, if it is appropriate to use a pie chart or mosaic plot to visualize a dataset, then it should also be appropriate to use a cartogram. You can refer to the examples of appropriate and inapropriate datasets for use with cartograms below for more guidance:
| Appropriate | Inapropriate |
|---|---|
| Total GDP by region | Human Development Index (or similar composite statistics) |
| Total population by region | Population density by region |
| Number of college gradates by region | Percentage of population that are college graduates by region |
For some data, the algorithm may not be able to compute a solution. For more information, please see our answer to “What went wrong?”
Please upload your boundary file with valid geometries. Ensure there are no self-intersections; you can use tools like Mapshaper to check and resolve such issues. After uploading, the system will extract properties from the boundary file. Only properties with unique values can be used as region names. For example, in a world map where 'country_names' are used as region names, Indonesia—an archipelago with many separate islands—should be represented as a 'MultiPolygon' rather than multiple 'Polygon' objects. Only when the region names are identified can go-cart.io generate a spreadsheet with numeric data for each region.
After defining your map, go-cart.io will generate a spreadsheet for filling in the data. You have two options to fill or modify the data:
In either method, you'll see the following columns:
Important Notes:
You can select a color scheme from the left panel under "Color." Each region will automatically be assigned a color based on the chosen scheme. Alternatively, choose "Custom" and click on a box in the "Color" column to select a color from the pop-up panel.
If you prefer to edit on your device, download the CSV file and use the hexadecimal color code format in the "Color" column. Learn more about this format here.
If you haven"t selected custom colors before downloading, you"ll find a "ColorGroup" column instead of "Color." Regions with the same "ColorGroup" will have the same color based on the chosen scheme. You can edit values in the "ColorGroup" as needed. Note that if the "Color" column is not empty, the visualization will use the "Color" column and ignore "ColorGroup."
An inset helps display regions that are geographically separated from the main area, like islands. Regions with different insets are processed separately and then combined based on the position defined in the "Inset" column.
We support 5 positions for insets: "C" (center), "L" (left), "R" (right), "T" (top), and "D" (down). If no position is entered, the default is "C."
It usually takes just a few seconds to generate one cartogram using this website. It will take longer if you have multiple data columns. A progress bar will give you an indication of how far the calculation has already proceeded. If you can’t see a cartogram at the end, please read our answer to “what went wrong?”
This website uses the Fast Flow-based method developed by Michael T. Gastner, Vivien Seguy, and Pratyush More. You can learn more about how this method works by reading the paper they published in the journal PNAS. If you publish an image produced by go-cart.io, please include a reference to the PNAS paper.
Have you tried to make a small region very large? For example, did you set the objective area of the Vatican City as large as the objective area of Russia? Or did you give a large region an objective area of zero? If yes, the algorithm will struggle with the severe distortions that would be necessary to show the data as a cartogram. In that case, a cartogram may not be the best way to represent your data. Please also see our answer to “What kind of data are suited for generating cartograms?”