Tessevolve (tesseract + evolve) is a platform for 2D, 3D, and 4D data visualization and exploration.
To try it out, click the Settings button and select some data to draw.
NOTE: Rendering takes a few seconds. If you are viewing in 4D, use your mouse wheel to scroll through the additional dimensions.
For more detailed info, click the FAQ button to the right.
Click the View button to center the visualization in your web browser.
If you have a compatible device (e.g. Oculus VR headset), you can then click the VR button in the bottom right to enter VR interaction mode. If you are having trouble with frame rate rendering/stuttering, turn off the lineage viewer.
Spheres represent points on the fitness landscape. They are evenly spaced and discrete even though the landscape is continuous so that there is space to see between them when you are inside the visualization. The color of a sphere represents its fitness, with warmer colors representing higher fitness. All colors are scaled to the viridis plasma color scale.
Boxes represent the birth of a new lineage. In this case, there are many new lineages because mutation rates are arbitrarily high. Like the points on the landscape, colors represent that lineage's fitness.
Lines represent sequential lineages. These are here as a visual aid to see how the lineage traverses the landscape.
Have fun! If you have questions or just want to talk about data visualization in VR, feel free to contact me!This is a visualization of evolutionary fitness landscapes. You can view them in 2D, 3D, and 4D (but you'll have to scroll through in 4D). The goal is for you to be able to understand and interpret 3D and 4D landscapes, which are difficult to show in flat images. I hope it helps!
Did you click the Settings button and select some data, then click Draw?
It takes a second to render, because A-Frame has to delete everything you see and put new stuff in its place. Also, you might need to back up (WASD or arrow keys) or turn around; the landscapes don't always appear in front of you. If you still see nothing, it's possible I accidentally broke something 😦. Please contact me at the link in the header if you keep having problems.
Unfortunately, there's not a great way to represent 4D data to our 3D brains. So what you're seeing is actually 3D slices of the complete 4D object. If you use your mouse wheel to scroll, the colors will change; this represents the fitness values for that 3D slice. Scrolling through multiple slices should give you some intuition about how the 4th dimension interacts with the other 3.
No, they aren't! The lines between nodes don't move, but the nodes themselves will change color as you scroll. If they are saturated with color, those nodes are on that slice. If they are greyed out and translucent, those nodes are on a different slice. Lines between nodes are displayed on all slices for simplicity and better rendering speed.
That's a great question! As far as I know, this is the first visualization using VR of either 3D or 4D fitness landscapes, so it certainly isn't perfect. But we would love to push the envelope even further. One thing we want to do next is add audio input with the VR headset's built-in directional sound capabilities. This might add another dimension, or might be a substitute for color for colorblind, low vision, or blind users of the software. If you are interested in working with me on that extension, PLEASE contact me, particularly if you fit one of the descriptors above.
The 2021 GECCO niching competition, with modifications by me to integrate the C++ versions of the landscapes into our digital evolution software and extend the landscapes you see here into 4D.
For full details, see the GitHub repo for this site. In summary:
MABE2 (pre-alpha): a heavily in-development digital evolution framework to evolve the phylogenies.
A-Frame (v1.2.0): an open source and shockingly straightforward web framework for virtual reality.
D3.js (v7): an open source, easy-to-use JavaScript library for data-driven HTML, which was absolutely invaluable for programatically creating the landscapes you see.
Bootstrap (v5): an open source web framework to connect the front-end to the back-end, which I used to make this website.
Open Iconic (v1.1.0): open source icons which I used for the buttons on this site.
And my laptop running EndeavourOS.
This work builds directly on Emily Dolson's previous work on visualizing 2D landscapes in 3D; the template for this website comes from her demo site for that paper. I'm additionally thankful to Charles Ofria, Clifford Bohm, and Vincent Ragusa for helpful discussions on this project.
I'm a PhD student creating methods and metrics to understand evolution at different scales. To learn more about me, see my website.