[Nauty] performance question

Adolfo Piperno piperno at di.uniroma1.it
Tue Feb 26 04:11:21 AEDT 2019


Hi Johannes

> Il giorno 25 feb 2019, alle ore 16:36, J.K.R. Bausch <jkrb2 at cam.ac.uk> ha scritto:
> 
> Hey everyone,
> 
> quick question. On http://pallini.di.uniroma1.it under "Experiments" the general performance appears to be such that a graph of <100 vertices takes from 1e-6 to 1e-2 ms, depending on the family of course. As a sanity check I wanted to compare this to my setup:

in all performance figures, time is in secs, not ms. By the way, I attach what I get 
on my mac with an Apple LLVM compiler (O3 flag on, 3,1 GHz Intel Core i7).
Best
- A

nauty27rc2 17:36 $ ./dreadnaut
Dreadnaut version 2.7 (64 bits).
> AS
> <G/G17
> t
  1 :  2 3 4 5 6 7 8 9 10 11 12;
  2 :  1;
  3 :  1;
  4 :  1;
  5 :  1;
  6 :  1;
  7 :  1;
  8 :  1;
  9 :  1;
 10 :  1;
 11 :  1;
 12 :  1 13 14 15 16 17;
 13 :  12;
 14 :  12;
 15 :  12;
 16 :  12;
 17 :  12;
> M500000
> -acx
4 orbits; grpsize=435456000; 13 gens; 105 nodes; maxlev=14
canupdates=1; cpu time = 0.0000127 seconds         <—— SPARSE NAUTY
> AT+
> x
4 orbits; grpsize=435456000; 4 gens; 1 node (1 peak); maxlev=0
canupdates=0; cpu time = 0.0000096 seconds      <—— TRACES
> AD+
> x
4 orbits; grpsize=435456000; 13 gens; 105 nodes; maxlev=14
canupdates=1; cpu time = 0.0000144 seconds       <—— DENSE NAUTY



------------------------------------
Adolfo Piperno
Dipartimento di Informatica
Sapienza Università di Roma
tel: +39 06 49918514
web: pallini.di.uniroma1.it




More information about the Nauty mailing list