[PPL-devel] A comparison with CDD and LSR
Roberto Bagnara
bagnara at cs.unipr.it
Mon Jul 12 10:23:04 CEST 2004
Hi there,
I have added a column to the previous table I sent, where I compare
the PPL both with CDD and the version 0.42 (the latest available one)
of LSR (http://cgm.cs.mcgill.ca/~avis/C/lrs.html) on the same benchmark
suite (the one of CDD).
Notice the "(*)" signs. These are meant to recall that LRS solves
a different, easier, problem than the one solved by CDD and PPL:
vertex enumeration _without_ minimization (at least this is what
I understand is the meaning of the sentence inserted by LRS in
the output: "Note! Duplicate rays may be present"). A similar
sentence is not present in the benchmarks involving the facet
enumeration problem, so I guess minimization is guaranteed
in these cases.
The test environment is exactly as before.
I must say I am pretty satisfied about the stability of the PPL
on this set of benchmarks.
Benchmark CDD time (s) PPL time (s) LRS time (s)
--------------------------------------------------------------------
ccc6.ext 0.80 0.07 3.20
ccp6.ext 1.12 0.11 3.99
cross10.ine 11.89 0.17 >2h (*)
cross12.ine 213.41 2.46 >2h (*)
cross6.ine 0.05 0.01 0.09 (*)
cross8.ine 0.70 0.04 33.70 (*)
cube10.ine 0.33 0.09 0.03 (*)
cube12.ine 1.74 1.64 0.20 (*)
cube8.ine 0.06 0.02 0.01 (*)
cyclic14-8.ext 0.09 0.02 0.01
cyclic16-10.ext 0.29 0.07 0.04
hexocta.ine 0.04 0.01 0.01 (*)
icododeca_m.ine 0.05 0.03 0.01 (*)
integralpoints.ine 0.06 0.01 0.01 (*)
irbox200-4.ext 1.24 0.05 0.06
kkd27_5.ine 0.10 0.09 0.02 (*)
kkd38_6.ine 0.37 4.30 0.04 (*)
prodst62.ext >2h 117.23 >2h
project1.ine 0.05 0.01 0.01 (*)
project2.ine 0.57 0.03 0.15 (*)
project2res.ine 0.11 0.02 0.02 (*)
reg600-5_m.ext 159.26 8.02 33.78
rhomtria_m.ine 0.06 0.02 0.01 (*)
sampleh8.ine >2h 111.57 5.13 (*)
Cheers,
Roberto
--
Prof. Roberto Bagnara
Computer Science Group
Department of Mathematics, University of Parma, Italy
http://www.cs.unipr.it/~bagnara/
mailto:bagnara at cs.unipr.it
More information about the PPL-devel
mailing list