🚀 go-pugleaf

RetroBBS NetNews Server

Inspired by RockSolid Light RIP Retro Guy

Article View: comp.theory
Article #1933

Is this NP-complete?

#1933
From: Brendan McKay
Date: Sun, 10 Jan 1993 15:48
17 lines
1164 bytes
A generalised regular family (GRF) of sets is represented by a 
vector (bot;top[1],...,top[n]) where bot and each top[i] are sets, 
and each top[i] is a superset of bot.  The GRF consists of all  
sets X such that bot <= X <= top[i] for some i, where "<=" means 
"subset".  A _simple_ GRF is one with n=1.

INSTANCE:  A GRF F; and integer K >= 0.
QUESTION:  Is F equal to the disjoint union of at most K
           simple GRFs?

What is the complexity of this problem?  I suspect it is NP-complete
but can't rule out some simple dynamic program (or something).
This problem arose from a graph computation that runs faster
if GRFs can be quickly split into a small number of simple GRFs.

Thanks,
Brendan McKay.
Technical Details
Message-ID: <1ipgg2INNobm@tyl.anu.edu.au>
Path: arc1-ca.pugleaf.net!sparky!uunet!crdgw1!rpi!think.com!ames!pacbell.com!network.ucsd.edu!munnari.oz.au!manuel.anu.edu.au!dubhe.anu.edu.au!tyl.anu.edu.au!not-for-mail

Show Full Headers
Complete Article Headers
Path: arc1-ca.pugleaf.net!sparky!uunet!crdgw1!rpi!think.com!ames!pacbell.com!network.ucsd.edu!munnari.oz.au!manuel.anu.edu.au!dubhe.anu.edu.au!tyl.anu.edu.au!not-for-mail
From: bdm@cs.anu.edu.au (Brendan McKay)
Newsgroups: comp.theory
Subject: Is this NP-complete?
Message-ID: <1ipgg2INNobm@tyl.anu.edu.au>
Date: 10 Jan 93 15:48:18 GMT
Distribution: inet
Organization: Australian National University
Lines: 17
NNTP-Posting-Host: tyl.anu.edu.au