Services & Resources / Wolfram Forums
-----
 /
MathGroup Archive
2007
*January
*February
*March
*April
*May
*June
*July
*August
*September
*October
*Archive Index
*Ask about this page
*Print this page
*Give us feedback
*Sign up for the Wolfram Insider

MathGroup Archive 2007

[Date Index] [Thread Index] [Author Index]

Search the Archive

Re: label graph vertices

  • To: mathgroup at smc.vnet.net
  • Subject: [mg74487] Re: [mg74472] label graph vertices
  • From: "Chris Chiasson" <chris at chiasson.name>
  • Date: Fri, 23 Mar 2007 19:01:10 -0500 (EST)
  • References: <200703220617.BAA17276@smc.vnet.net>

This is possible. GraphPlot adds additional functionality to the
graphics function Text. Instead of a position, you can give an integer
that corresponds to a graph node number.

I used that functionality in this package:

http://bond-graphs.googlecode.com/svn/trunk/BondGraphs/BondGraphs.nb

Take a look at the definition of BondGraphPlot, specifically the
option for VertexStyle that is fed to GraphPlot.

I personally have a hard time making Mathematica 5.2 draw nice directed
graphs. It certainly isn't a replacement for Visio. On the other hand,
neither is Graphviz.
1. It is difficult to specify that the arrows between nodes should end
or begin at a non-circular a bounding box border.
2. It would be nice to see more options for the node layout algorithms
& edge routing.

On 3/22/07, alexxx.magni at gmail.com <alexxx.magni at gmail.com> wrote:
> greetings,
> after having used for some time a very powerful program (Graphviz) for
> displaying the graph structures I'm working on, I decided to give it a
> try under Mathematica, since I use M. for many other related tasks.
>
> I started from DiscreteMath`GraphPlot`, but from what I can see
> everything is done using an adjacency matrix composed of (0,1)'s,
> right?
> That is, my problem is that I need to LABEL the vertices with a name,
> instead of a position.
> >From http://documents.wolfram.com/mathematica/functions/AdvancedDocumentationGraphPlot
> I thought for a moment it was possible (e.g. see graph under
> "MaximalBipartiteMatching" with Tom,Rob,Adam etc), but I then realized
> it was demonstration-only, and no code is present to do it
>
> am I right, or there is a way?
>
> thanks for any help...
>
> Alessandro Magni
>
>
>


-- 
http://chris.chiasson.name/


  • Prev by Date: Re: About function to evaluation
  • Next by Date: Re: Advice for writing small Scheme interpreter
  • Previous by thread: label graph vertices
  • Next by thread: Re: label graph vertices