[SCIP] problems with "infinity"

[Marc Pfetsch]

Hi James,

thanks for pointing this out. We need to discuss how to take care of 
infinite coefficients in linear constraints or whether we want to forbid 
them.

Best

Marc


On 28/09/2023 17:07, James Cussens wrote:
> Hi,
> 
> I have a linear constraint:
> 
>   [linear] <scoreub>: <score>[C] -1e+20<ch#0>[B] -1e+20<ch#1>[B] 
> -1e+20<ch#2>[B] -1e+20<ch#3>[B] -1e+20<ch#4>[B] -1e+20<ch#5>[B] 
> -1e+20<ch#6>[B] <= 0;
> 
> which (in debug mode) is causing the following problem:
> 
> gobnilp: src/scip/set.c:6168: SCIPsetIsEQ: Assertion 
> `((!SCIPsetIsInfinity(set, val1) || !SCIPsetIsInfinity(set, val2)) && 
> (!SCIPsetIsInfinity(set, -val1) || !SCIPsetIsInfinity(set, -val2))) || 
> val1 == val2' failed.
> 
> Thread 1 "gobnilp" received signal SIGABRT, Aborted.
> __pthread_kill_implementation (no_tid=0, signo=6, 
> threadid=140737350973376) at ./nptl/pthread_kill.c:44
> 44    ./nptl/pthread_kill.c: No such file or directory.
> (gdb) bt
> #0  __pthread_kill_implementation (no_tid=0, signo=6, 
> threadid=140737350973376) at ./nptl/pthread_kill.c:44
> #1  __pthread_kill_internal (signo=6, threadid=140737350973376) at 
> ./nptl/pthread_kill.c:78
> #2  __GI___pthread_kill (threadid=140737350973376, signo=signo at entry=6) 
> at ./nptl/pthread_kill.c:89
> #3  0x00007ffff6a42476 in __GI_raise (sig=sig at entry=6) at 
> ../sysdeps/posix/raise.c:26
> #4  0x00007ffff6a287f3 in __GI_abort () at ./stdlib/abort.c:79
> #5  0x00007ffff6a2871b in __assert_fail_base (fmt=0x7ffff6bdd150 
> "%s%s%s:%u: %s%sAssertion `%s' failed.\n%n",
>      assertion=0x555559676ce0 "((!SCIPsetIsInfinity(set, val1) || 
> !SCIPsetIsInfinity(set, val2)) && (!SCIPsetIsInfinity(set, -val1) || 
> !SCIPsetIsInfinity(set, -val2))) || val1 == val2", file=0x55555966f220 
> "src/scip/set.c", line=6168, function=<optimised out>) at 
> ./assert/assert.c:92
> #6  0x00007ffff6a39e96 in __GI___assert_fail (
>      assertion=0x555559676ce0 "((!SCIPsetIsInfinity(set, val1) || 
> !SCIPsetIsInfinity(set, val2)) && (!SCIPsetIsInfinity(set, -val1) || 
> !SCIPsetIsInfinity(set, -val2))) || val1 == val2", file=0x55555966f220 
> "src/scip/set.c", line=6168, function=0x555559678098 
> <__PRETTY_FUNCTION__.72> "SCIPsetIsEQ")
>      at ./assert/assert.c:101
> #7  0x0000555557a873d8 in SCIPsetIsEQ (set=0x55555b42fc70, 
> val1=-2.6911787705190607e+22, val2=-1e+20) at src/scip/set.c:6168
> #8  0x00005555578f6775 in SCIPisEQ (scip=0x55555b28f340, 
> val1=-2.6911787705190607e+22, val2=-1e+20) at src/scip/scip_numerics.c:454
> #9  0x0000555556e3fe2f in checkEqualObjective (scip=0x55555b28f340, 
> consdata=0x55555b3fd7f8, scale=0x7fffffffaad8, offset=0x7fffffffaad0)
>      at src/scip/cons_linear.c:10322
> 
> so it looks like the problem is caused by checkEqualObjective generating 
> a value ( -2.6911787705190607e+22 ) which is less than -1+e20 (ie less 
> than "negative infinity"). I can easily get round this (and indeed 
> should do since constraints with infinite coefficients aren't a great 
> idea!) but it might be worth adding some sort of check for "infinite" 
> coefficients in linear constraints.
> 
> James
> 
> James Cussens
> Room MVB 3.26
> School of Computer Science, University of Bristol
> Phone: +44 (0)117 455 8723
> https://jcussens.github.io/ <https://jcussens.github.io/>
> Funded PhDs available in Bristol in the following areas: Data Science 
> <http://www.bristol.ac.uk/cdt/compass/>, Interactive AI 
> <http://www.bristol.ac.uk/cdt/interactive-ai/>, Cyber Security 
> <http://www.bristol.ac.uk/cdt/cyber-security/> or Digital Health 
> <http://www.bristol.ac.uk/cdt/digital-health/>.
> 
> _______________________________________________
> Scip mailing list
> Scip at zib.de
> https://listserv.zib.de/mailman/listinfo/scip