[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