It is a great joy of mine to make mathematical diagrams and art — from impressing students with quick sketches of three-dimensional surfaces, to helping colleagues by illustrating graphs for their research, to formalizing collaborators' and my own ideas by translating them into a visual form.
Since early 2021, I have had the pleasure of studying De Bruijn sequences, universal partial cycles, and perfect necklaces, all of which can be represented as Hamiltonian or Eulerian circuits on different dimensions of De Bruijn graphs, or the more general astute graphs (also known as Praeger Xu graphs). In fact, my collaborators and I have a submitted paper on this work, available to view on ArXiv here: https://arxiv.org/abs/2310.13067.
Below are some of the hand-drawn (and occasionally LaTeχ-coded) graphs that I have created over the course of my research:
B(2,2): De Bruijn graph for a=2, n=2.
B(2,3): De Bruijn graph for a=2, n=3.
Eulerian circuits of B(2,3) up to symmetries.
Assorted De Bruijn graphs (sorry, notation has fluctuated over time).
Hamilton cycle and Euler tour representations of the De Bruijn cycle (0010110000111101).
(Generalized) Hamilton and Euler representations of the universal partial cycle (001⟡110⟡).
(Generalized) Hamilton and Euler representations, on DB(4,4) and DB(4,3) respectively, of the universal partial cycle
(001⟡110⟡003⟡112⟡021⟡130⟡023⟡132⟡201⟡310⟡203⟡312⟡221⟡330⟡223⟡332⟡).
Okay, technically the next graphs are only hypothetically maybe subgraphs of De Bruijn graphs — if you can show they are, then get back to me! 😄
The structure of the (generalized) Hamilton representation of a universal partial cycle with any alphabet size and word length is similar to the structure of the astute graph with parameters relating to the diamondicity of the upcycle.
*actually I noticed a minor error in the last image here, that what's marked as n is really n-1, but who's counting. ...it's me. I'm counting.
I learned of tri- and hexa-hexaflexagons from Vihart in 2012, and since then I have been fascinated by them, to the extent that I typically keep a hexa-hexaflexagon in my wallet for just such an occasion as presented itself recently when I had the pleasure of introducing a group of combinatorists to the beautiful structure of this novel object and its state diagram. During that discussion, it occurred to me that I hadn't ever tried to construct a higher than six-faced version of this object (I had made many tri- and hexa-hexaflexigons, and even some tetraflexagons). I was very pleased then when my first attempt at construction indeed resulted in a dodeca-hexaflexagon, whose state diagram is most satisfyingly exactly what I would have hoped.
Tri-hexaflexagon state diagram...
Hexa-hexaflexagon state diagram...
Dodeca-hexaflexagon state diagram!
I find myself tempted to construct an icositetra-hexaflexagon... Would its state diagram follow the pattern too?
...only one way to find out. 😉