Re: [Bug-glpk] Documentation: RHS for minimum cost flow

[Robbie Morrison]

Hello Andrew, Xypron

------------------------------------------------------------
To:           glpk xypron <address@hidden>
Subject:      Re: [Bug-glpk] Documentation: RHS for minimum cost flow
Message-ID:  <address@hidden>
From:         Andrew Makhorin <address@hidden>
Date:         Wed, 18 Apr 2012 11:46:31 +0400
------------------------------------------------------------

>> for minimum cost flow problems the inflow to the
>> nodes is significant.  It is specified by parameter
>> RHS of the vertices.
>>
>> My understanding is that a positive value of RHS is a
>> flow to the node and a negative value of RHS is a
>> flow out of the node.
>
> Exactly the opposite: a positive rhs value is the
> supply at the node, and negative rhs value is the
> demand at the node. This is a standard convention.

That is not "exactly the opposite" -- that is the
same.  Xypron and Andrew are following the same
convention.

>> I was unable to find a definition in file glpk-4.47/tables.pdf.
>>
>> Could you, please, add the missing information.
>
> Please see Section 2.1 "Minimum cost flow problem",
> Subsection 2.1.1 "Background" on pp.24-25 (file
> graphs.pdf).

The official explanation looks good to me.

cheers, Robbie
---
Robbie Morrison
PhD student -- policy-oriented energy system simulation
Technical University of Berlin (TU-Berlin), Germany
University email (redirected) : address@hidden
Webmail (preferred)           : address@hidden
[from Webmail client]