Creating Better Cartograms with New Techniques
Learn how new approaches improve cartograms for visual data representation.
― 5 min read
Table of Contents
- The Challenge
- The Solution: Meshes and Numbers
- The Two Options: On a Flat World or Round World
- Flat and Floppy
- Round and Ready
- The Best of Both Worlds
- Comparing Methods: A Shape Showdown
- Rubber Bands and Fluids
- Keeping Shapes Intact
- The Tradeoff: Time vs. Accuracy
- How do We Create Our Cartograms?
- Step 1: Start with a Triangular Mesh
- Step 2: Adjust Vertices
- Step 3: Check the Data
- Final Steps and Cartogram Creation
- The Results: Liquid Earth
- Why It Matters
- The Future of Cartograms
- Conclusion
- Original Source
- Reference Links
Cartograms are special maps that change the size of countries or areas based on Data, like population or income. Instead of showing the actual size of places, they reshape them to visually represent information. If you think of it like a game of Tetris, fitting pieces together, that's a bit like what cartograms do-except they sometimes make the pieces look a little weird!
The Challenge
Creating cartograms can sometimes lead to problems. When we change the shapes of places to fit the data, we often end up with Distortions. This means that places can look like they are stretched, squished, or warped out of shape-kind of like a funhouse mirror! So, how can we create cartograms that keep the information intact while not making the shapes look like they belong in a circus?
Meshes and Numbers
The Solution:We came up with a new way to create cartograms by using something called meshes. Imagine a mesh as a web of Triangles that covers the globe. With the help of some smart math and numbers, we can adjust these triangles to minimize how much distortion happens.
The Two Options: On a Flat World or Round World
We have two main ways to work on our cartograms: we can treat the world as flat, like a piece of paper, or round, like a ball. Each way has its pluses and minuses.
Flat and Floppy
In a flat world, we take the globe and squish it down. This is like taking a basketball and making it into a pancake. We place triangles on the flattened version and use our numerical magic to make sure that the areas on the map match our data.
Round and Ready
The round approach is more like playing with a globe. We keep everything in its three-dimensional shape and only change the details as needed. This cuts down on the weirdness that can come with flattening everything out. After we adjust the triangles, we project them back onto our flat maps.
The Best of Both Worlds
We also have a hybrid option, which combines both methods. This is like making a smoothie with fruits from both the flat and round worlds. We optimize the triangles in our round view while keeping in mind how they will look when we finally flatten everything out. The result? Cartograms that keep their shapes while showing the data clearly.
Comparing Methods: A Shape Showdown
To see how our new method stacks up, we compared it to older methods. Some older techniques can make maps look a bit too distorted. For instance, imagine trying to stretch a rubber band to fit a shape. You might succeed, but it can look a bit lumpy!
Rubber Bands and Fluids
Two older methods are the rubber sheet method and the diffusion method. The rubber sheet method stretches and shrinks parts of the map until they fit. The diffusion method imagines the map as a sponge soaking up information, spreading out evenly. While both can create useful maps, they can also introduce a lot of distortion.
Keeping Shapes Intact
With our new approach, we focused on keeping the shapes as accurate as possible. For example, when adjusting smaller islands or countries, we find ways to stretch them perfectly without ruining their shape. No more squished states or floppy islands!
The Tradeoff: Time vs. Accuracy
One downside to our method is that it requires more time to compute than the old methods. While other techniques produce results in minutes, ours might take hours. Think of it like cooking a perfect sauce: it takes time and patience for the best flavor!
How do We Create Our Cartograms?
Let’s break down how we actually create these cartograms step-by-step.
Step 1: Start with a Triangular Mesh
We begin with a triangular mesh covering the globe. Imagine a soccer ball, but instead of a smooth surface, it’s made of tiny triangle pieces.
Step 2: Adjust Vertices
Next, we move the “points” of these triangles (called vertices) to try and get them to match the desired shapes based on our data. This adjustment is where the numerical optimization comes in. We carefully calculate how to shift each point to minimize distortion.
Step 3: Check the Data
Throughout the process, we keep checking to make sure the areas on our map still reflect the data we want to show. It’s like going back and checking your math homework to ensure you got all the answers right!
Final Steps and Cartogram Creation
Once we’ve adjusted the mesh, we create the final cartogram. We project the shapes back to a flat map, ensuring we don’t introduce unwanted distortion.
The Results: Liquid Earth
One of the exciting outcomes of our method is the Liquid Earth projection. This cartogram looks like a beautifully distorted world map but maintains the right area sizes for different regions. If Earth were a liquid, this is how it might look!
Why It Matters
Improving cartogram accuracy matters for many reasons. Better maps can lead to better understanding and communication of important data. Whether it’s population density, economic data, or other factors, a clear visual representation helps everyone from students to policymakers.
The Future of Cartograms
As we move forward, the goal is to further improve these methods. We want to find ways to make them faster and easier to use while still delivering high-quality results. It would be like upgrading from a flip phone to a smartphone-same purpose, but much better!
Conclusion
In wrapping all this up, creating cartograms with a focus on reducing distortion is like learning the best dance moves for a wedding. It takes practice, timing, and a bit of creativity to get it right. With the new mesh and optimization methods, we can create cartograms that are both informative and visually appealing, helping everyone better understand the beautiful world we live in.
Title: Minimum-distortion continuous cartograms by numerically optimized meshes
Abstract: We present an algorithm for creating contiguous cartograms using meshes. We use numerical optimization to minimize cartographic error and distortion by transforming the mesh vertices. The vertices can either be optimized in the plane or optimized on the unit sphere and subsequently projected to the plane. We also present a hybrid "best of both worlds" method, where the vertices are optimized on the sphere while anticipating the distortion caused by the final projection to the plane. We show a significant improvement in the preservation of region shapes compared to existing automated methods. Outside the realm of cartograms, we apply this hybrid technique to optimized map projections, creating the Liquid Earth projection.
Last Update: Nov 26, 2024
Language: English
Source URL: https://arxiv.org/abs/2411.17129
Source PDF: https://arxiv.org/pdf/2411.17129
Licence: https://creativecommons.org/licenses/by/4.0/
Changes: This summary was created with assistance from AI and may have inaccuracies. For accurate information, please refer to the original source documents linked here.
Thank you to arxiv for use of its open access interoperability.
Reference Links
- https://github.com/rsargentmath/cartogram-gradient-descent
- https://doi.org/10.2140/pjm.1966.16.1
- https://doi.org/10.1111/j.0033-0124.1985.00075.x
- https://doi.org/10.1073/pnas.0400280101
- https://doi.org/10.1073/pnas.1712674115
- https://doi.org/10.1111/cgf.12647
- https://doi.org/10.1080/15230406.2016.1270775
- https://doi.org/10.1080/23729333.2020.1824174
- https://kunimune.blog/2023/12/29/introducing-the-elastic-projections/
- https://doi.org/10.1080/15230406.2017.1408033
- https://graphallthethings.com/posts/map-projections-2
- https://doi.org/10.1090/S0025-5718-1980-0572855-7
- https://doi.org/10.1111/cgf.12932
- https://rsargentmath.github.io/posts/liquid_earth/
- https://doi.org/10.1080/13658816.2012.709247
- https://doi.org/10.1080/00330124.2011.639613
- https://www.sunsp.net/download.html