Skip to content

Edge-chromatic 3-critical graphs — census data

This site documents the public research data repository for Le Chen and Songling Shan, Structure and census of edge-chromatic 3-critical graphs (2026). The repository contains census outputs, verification code, and audit reports supporting the paper's computational classification statements.

A nontrivial edge-chromatic 3-critical graph here means a graph that is connected, has maximum degree 3, has chromatic index 4, becomes 3-edge-colorable after deleting any edge, and has no 3-overfull subgraph.

Hash source of truth

The canonical SHA-256 values for census files live in the repository README.md. This site deliberately does not duplicate those hashes; CI checks the files against the README table.

Census results

The table records the number of nontrivial survivors for each order in the public census range.

Order Nontrivial survivors Status
4 0 determined
5 0 determined
6 0 determined
7 0 determined
8 0 determined
9 1 determined
10 0 determined
11 2 determined
12 0 determined
13 14 determined here
14 0 determined
15 94 determined here
16 0 determined
17 774 determined here
18 0 determined
19 6,984 determined here
20 0 determined
21 70,530 determined here
22 1 Brinkmann--Steffen survivor reproduced
23 782,186 determined here
24 9 Brinkmann--Steffen count reproduced

Even orders below 22 have no nontrivial survivors. The nontrivial-survivor counts for all odd orders through 23 are recorded here. The order-22 and order-24 computations reproduce the Brinkmann--Steffen counts; eight of the nine order-24 survivors are triangle blowups of the unique order-22 survivor, and the remaining order-24 graph is triangle-free.

Repository contents

  • results/ — per-order survivor census files named order_N_delta_3.json or order_N_delta_3.json.gz.
  • reports/ — audit JSON files recording independent checks and pipeline summaries.
  • code/ — the census pipeline plus classification and audit scripts, including code/scripts/classify_census_characterization.py (reproduces the paper's categorization table) and code/scripts/check_hashes.py (verifies the census SHA-256 hashes).
  • README.md — canonical census-file SHA-256 table and repository overview.