Sign in

Patrick Schnider

@schnpatr.bsky.social
108 followers 80 following 175 posts

Mathematician/Computer Scientist interested in discrete and computational geometry and topology. Working at University of Basel and ETH Zürich. people.inf.ethz.ch/schnpatr

PostsRepliesMedia
Patrick Schnider @schnpatr.bsky.social · 07/10/2026
3b. It also does not provide adequate context, neither does it mention the natural new interesting questions this result raises (for example about crossing families in simple topological drawings).
010
Patrick Schnider @schnpatr.bsky.social · 07/10/2026
3a. it is noticeable that the paper was not written by a person normally working in discrete geometry. The write-up, while being precise, puts a lot of focus on standard arguments and does not highlight which ideas are genuinly new.
120
Patrick Schnider @schnpatr.bsky.social · 07/10/2026
2. from what I see, the techniques immediately generalize to generalized configurations (aka rank 3 oriented matroids aka pseudolinear drawings). I have not fully checked this yet though.
100
Patrick Schnider @schnpatr.bsky.social · 07/10/2026
1. the proof seems to be correct and the new argument is elegant and quite simple. The underlying heavy machinery was already developed pre-AI, it turned out that we just needed one more nice idea.
100
Patrick Schnider @schnpatr.bsky.social · 07/10/2026
Given a set P of n points in the plane, no three collinear, how many pairwise crossing segments with disjoint endpoints in P are there? It was conjectured that this number should always be linear in n, and we now finally have a proof for it. Having read through the paper, here are some thoughts:
100
Patrick Schnider @schnpatr.bsky.social · 07/10/2026
After taking a summer break from social media, I decided to make a post about math again. With another 700ish AI results announced today I want to talk about another problem close to my heart which Tim Gehrunger recently solved using AI: crossing families arxiv.org/abs/2609.29449
arxiv.org
Linear crossing families via dual levels
Building on the work of Pach, Rubin, and Tardos, we prove that every set of $n\ge2$ points in the real plane with no three collinear contains at least $cn$ pairwise crossing segments with distinct end...
110
Patrick Schnider @schnpatr.bsky.social · 11/07/2026
Spotted at the University of Vienna. Very impressive list from a perspective of discrete geometry, with the names of two of the most fundamental theorems (Radon and Helly) appearing.
020
Patrick Schnider @schnpatr.bsky.social · 10/07/2026
Great talk by Hana Dal Poz Kourimska on how to smooth a manifold, with amazing slides.
010
Patrick Schnider @schnpatr.bsky.social · 10/07/2026
Facundo Memoli analyed the Gromov-Hausdorff distance between spheres. Some cool quantitative version of Borsuk-Ulam popping up.
110
Patrick Schnider @schnpatr.bsky.social · 10/07/2026
Herbert Edelbrunner talked about a connection between Euclidean Minimum Dpanning Trees and chromatic persistent homology.
110
Patrick Schnider @schnpatr.bsky.social · 10/07/2026
Day 2 of #FoCM26. A nice plenary talk by Beatrice Acciaio about the space of models for discrete stochastic processes. I don‘t know a lot about this, but still found the examples very interesting.
110
Patrick Schnider @schnpatr.bsky.social · 09/07/2026
Right after my talk there was also a poster session with some really interesting works. My personal highlight was Jake Cordes‘ poster on an application of persistent homology in climate science.
010
Patrick Schnider @schnpatr.bsky.social · 09/07/2026
Chiara Meroni talked about directional convexity, a very interesting notion of convexity where surprisingly many natural things don‘t work anymore. It is even conjectured that directional convex hulls are in general not semialgebraic.
110
Patrick Schnider @schnpatr.bsky.social · 09/07/2026
Other highlights of day 1 of #FoCM26: the second plenary talk was by Arno Kuilaars on (among other things) random tilings of hexagons, which lead to some fascinating effects and very pretty pictures.
110
Patrick Schnider @schnpatr.bsky.social · 09/07/2026
The conference started with an amazing talk by Uli Wagner on face numbers of polytopes and arrangements
010
Patrick Schnider @schnpatr.bsky.social · 09/07/2026
Spending a few days in Vienna at #FoCM26. Looking forward to a hopefully very interesting conference.
120
Patrick Schnider @schnpatr.bsky.social · 03/07/2026
I recently learned that my research was featured in an article in the Süddeutsche Zeitung on math for cutting pizzas www.sueddeutsche.de/projekte/art... It‘s unfortunately paywalled, if you want a copy dm me :)
sueddeutsche.de
Wie man mit Mathematik eine Pizza aufteilt
So teilt man Essen gerecht, auch wenn die Esser unterschiedliche Dinge mögen – oder im Hyperraum unterwegs sind.
042
Patrick Schnider @schnpatr.bsky.social · 29/06/2026
Welcome to the department @christophepiveteau.bsky.social!
110
Patrick Schnider @schnpatr.bsky.social · 24/06/2026
Spending a week in beautiful Wergenstein for our workshop „Combinatorial, Topological and Algorithmic Aspects of Tverberg‘s Theorem“
010
Patrick Schnider @schnpatr.bsky.social · 15/06/2026
So far, the conjecture had been confirmed in all dimensions that are a power of 2. Given this counterexample for the general conjecture, could it be that nd masses can be bisected with n hyperplanes only if d is a power of 2 (or if n=1, where this is just the Ham Sandwich theorem)?
000
Patrick Schnider @schnpatr.bsky.social · 15/06/2026
Progress on one of my favorite problems: Langerman‘s conjecture states that any nd mass distributions in d-space can be simultaneously bisected by n hyperplanes. Pablo Soberon just informed me that he and Alfredo Hubard found 6 masses in 3-space that cannot be bisected by 2 planes!
100
Reposted by Patrick Schnider
Paul Goldberg @paulwgoldberg.bsky.social · 11/06/2026
5-year postdoctoral research fellowship in theoretical Computer Science (also 4 other research fields), a nice opportunity! Needs PhD to be awarded between 1/Aug/24 and 1/Oct/27. Please share with whoever may be suitable! www.asc.ox.ac.uk/post-doctora...
asc.ox.ac.uk
Post-Doctoral Research Fellowships | All Souls College
088
Patrick Schnider @schnpatr.bsky.social · 11/06/2026
A question that stomped quite a few of my students: consider an naive version of the union find data structure (that is, without ranking or path compression). Given elements 0,1,2,3,4,5, what is a sequence of calls to union such that after the calls the representative tree is a path with root 0?
000
Reposted by Patrick Schnider
Clément Canonne @ccanonne.github.io · 26/05/2026
"I haven't really seen further than others, but they were looking in another direction." Academic tip: we're not all Newtons, and that's completely OK! You don't need to be a genius to do good research.
1234
Patrick Schnider @schnpatr.bsky.social · 18/05/2026
A few weeks ago I gave a talk at the New York Geometry Seminar about the alpha-Ham-Sandwich theorem. The talk is now online on YouTube: m.youtube.com/watch?v=MEiq...
m.youtube.com
Patrick Schnider: The alpha-Ham-Sandwich Theorem revisited
YouTube video by Computational Geometry
010
Patrick Schnider @schnpatr.bsky.social · 29/04/2026
If you want to know a bit more about my recent paper with Juan, Ernst and Bastian, watch this video of Juan explaining it :)
021
Patrick Schnider @schnpatr.bsky.social · 28/04/2026
A new paper on the #arXiv: with Daniel, Henry, Fabian, Irene and Birgit we study piercing double wedges (which in the primal corresponds to finding lines intersecting some segments and avoiding some others). We give algorithms but also Gallai-type combinatorial results. arxiv.org/abs/2604.23330
arxiv.org
Bowties and Hourglasses: Intersections of Double-Wedges (or Stabbing and Avoiding Line Segments)
We study the common intersection of arrangements of double-wedges. We consider arrangements where double-wedges may be either bowties (which do not contain a vertical line) or hourglasses (which conta...
000
Reposted by Patrick Schnider
DMI | University of Basel @dmi.unibas.ch · 28/04/2026
Prof. Susanna Zimmermann has been invited to speak at the International Congress of Mathematicians — a major honour. First woman from the University of Basel. A specialist in algebraic geometry. dmi.unibas.ch/en/news-even... #ICM2026 #Mathematics #UniBasel
S. Zimmermann in Basel
021
Patrick Schnider @schnpatr.bsky.social · 28/04/2026
A new paper of mine finally on the arXiv :)
000
Patrick Schnider @schnpatr.bsky.social · 17/04/2026
A special treat for the students in my course „Introduction to Topological Data Analysis“: Bastian Rieck (@pseudomanifold.topology.rocks) gave a very interesting guest lecture on topological methods in deep learning.
121
Patrick Schnider @schnpatr.bsky.social · 09/04/2026
Very nice explanation of one of my favorite theorems in math, the Ham Sandwich theorem, by @manonym025.bsky.social www.scientificamerican.com/article/the-...
scientificamerican.com
The mathematically correct way to slice a pizza
The intermediate value theorem shows us how to find an even center on an irregular shape
010
Reposted by Patrick Schnider
Thatchaphol Saranurak @eigx.bsky.social · 06/04/2026
🔗 Full materials: sites.google.com/site/thsaran...
sites.google.com
Thatchaphol Saranurak - Graph Decomposition
What is this course about? This course studies modern graph algorithms through the lens of graph decomposition. Over the last two decades, expander decompositions and hierarchies have become some of ...
191
Patrick Schnider @schnpatr.bsky.social · 27/03/2026
Joseph Dorfer spoke about flipping spanning trees of convex point sets, showing that computing the flip distance is NP-hard. This paper gave him the crucial ideas for his recent breakthrough on NP-hardness of the flip distance of convex triangulations (arxiv.org/abs/2602.22874)
000
Patrick Schnider @schnpatr.bsky.social · 27/03/2026
ETH-Alumni Justin Dallant spoke about his new lower bounds for the number of triangulations parametrized by the convex hull size. First time that I heard about the intriguing unimodality conjecture.
100
Patrick Schnider @schnpatr.bsky.social · 27/03/2026
Day 3 of #EuroCG26 started with an invited talk by Jean Cardinal on compact representations of graphs, where he particularly focused on biclique covers. An interesting topic I did not know about before.
100
Patrick Schnider @schnpatr.bsky.social · 27/03/2026
My students Jette Gutzeit and Kalani Kistler talked about their joint work with Anna Schenfisch and Tim Ophelders on VPHT-reconstructible graphs, a project that started in our course „Projects in TDA“.
000
Patrick Schnider @schnpatr.bsky.social · 27/03/2026
Niko Fink presented our joint paper on „garment numbers“ (arxiv.org/abs/2603.05339). Given that we also try to find „bowties“ in point sets, he wore one for this occasion.
110
Patrick Schnider @schnpatr.bsky.social · 27/03/2026
Highlights from day 2 of #EuroCG26. The day started again with an invited talk, today by Maike Buchin who took us on a tour of 35 years of Frechét distance. A very interesting talk, and lots of great illustrations of dogs on walks.
100
Patrick Schnider @schnpatr.bsky.social · 26/03/2026
The day ended with a long business meeting (2h instead of the scheduled 1h). #EuroCG28 will take place in Passau, and authors will have the option of only publishing a 1-page abstract in the booklet of abstracts.
000
Patrick Schnider @schnpatr.bsky.social · 26/03/2026
Another highlight was the talk of Sarita de Berg on the contiguous art gallery problem. While the original art gallery problem is ETR-complete, this variant can be solved in polynomial time, and Sarita and her coauthors have developed an optimal algorithm for this.
100
Patrick Schnider @schnpatr.bsky.social · 26/03/2026
Some personal highlights from day 1 of #EuroCG26: the day started with a great invited talk by Marcus Schaefer about penny graphs and the existential theory of the reals. I recently worked quite a bit in ETR, and it was very interesting to hear one of the founding fathers.
120
Patrick Schnider @schnpatr.bsky.social · 26/03/2026
I am at #EuroCG26 in Hagen this week. It is once again an amazing conference with lots of interesting talks and discussions. There are also some of my MSc students from ETH here, attending their first ever conference, and it‘s great to see how they are welcomed by the CG community.
000
Patrick Schnider @schnpatr.bsky.social · 17/03/2026
Very cool question with an elegant solution that I also discussed last week in my lecture (I‘m not going to say which lecture so as not to give hints about the answer) ☺️
040
Patrick Schnider @schnpatr.bsky.social · 17/03/2026
There is also a fun ghost hunt in the castle, I‘m currently at 26 found ghosts.
000
Patrick Schnider @schnpatr.bsky.social · 17/03/2026
Spending an interesting week at Schloss Dagstuhl at the workshop „Intractability in Discrete Geometry and Topology“. A very nice venue, some cool open problems, and even cooler participants. www.dagstuhl.de/seminars/sem...
131
Patrick Schnider @schnpatr.bsky.social · 10/03/2026
Last semester, I organised „Projects in Topological Data Analysis“, where students from several universities collaborate on problems in TDA. My Students Jette and Kalani, together with their mentors Anna and Tim have just uploaded their findings from their project to ArXiv: arxiv.org/abs/2603.07809
arxiv.org
Which Vertical Graphs are Non VPHT Reconstructible?
The verbose persistent homology transform (VPHT) is a topological summary of shapes in Euclidean space. Assuming general position, the VPHT is injective, meaning shapes can be reconstructed using only...
000
Reposted by Patrick Schnider
cs-cg-arxiv.bsky.social @cs-cg-arxiv.bsky.social · 06/03/2026
📚 New in arXiv cs.CG: Garment numbers of bi-colored point sets in the plane
arxiv.org
Garment numbers of bi-colored point sets in the plane
New paper in cs.CG (Computational Geometry)
011
Patrick Schnider @schnpatr.bsky.social · 06/03/2026
We show that this statement is true for other structures on four points, which we give names like pants, skirts or cravates. This motivates the name „garment numbers“ in our title.
000
Patrick Schnider @schnpatr.bsky.social · 06/03/2026
Say you are given n points in the plane (no 3 on a line), colored red and blue. It is conjectured that if n is large enough you can always find a quadrilateral spanned by 4 points of the same color whose interior does not contain any other points.
100
Patrick Schnider @schnpatr.bsky.social · 06/03/2026
A new paper on the ArXiv. This one is a result of the research week on geometric graphs in Trier in August 2024. Together with my coauthors Oswin, Helena, Niko, Maarten and Pepa we study a variant of the empty monochromatic quadrilateral problem. arxiv.org/abs/2603.05339 This will be at #EuroCG26
arxiv.org
Garment numbers of bi-colored point sets in the plane
We consider colored variants of a class of geometric-combinatorial questions on $k$-gons and empty $k$-gons that have been started around 1935 by Erdős and Szekeres. In our setting we have $n$ points ...
100