DIR Return Create A Forum - Home
       ---------------------------------------------------------
       knight tours
  HTML https://knighttours.createaforum.com
       ---------------------------------------------------------
       *****************************************************
   DIR Return to: general Hamiltonian cycles problem
       *****************************************************
       #Post#: 14--------------------------------------------------
       Vandegriend/Culberson 's 1998 - program
       By: gsgs Date: June 22, 2023, 5:18 am
       ---------------------------------------------------------
       available here :
  HTML https://www.jair.org/index.php/jair/article/view/10213/24281
       paper with analysis :
  HTML https://scholar.archive.org/work/fbbnjecre5gzdj2mbivb5tpq3e/access/wayback/http://jair.org/media/512/live-512-1717-jair.pdf
       2021 tests + comparisons :
  HTML https://arxiv.org/pdf/2107.00314
       it's typically 10-100 times slower than Chalaturnyk's program,
       except for the triangle-problem and forced-edge problem,
       which it also has, but often less severe.
       It also benefits from small-border-reordering and restarts
       and solves 392 out of 638 FHCP-challenge-graphs
       It cannot count all HCs in a graph and only reads one graph at a
       time,
       but has other features, ,e.g, generating and testing random
       graphs.
       with <3600 vertices and max.degree<50 in 5 hours
       with timelimit=10s per test.
       I get segmentation errors for bigger graphs.
       The maximum number of vertices is 1600 by default,
       but I could change it to 3600 in the code.
       *****************************************************