Re: [igraph] igraph-help Digest, Vol 79, Issue 19

[Bita Shams]

Special Thanks For your comprehensive explanation.

I also want to know is there any way to obtain the exponent element of graphs generated by Barabasi model or not?

Best Regards,

B.S.

---

Date: Wed, 20 Feb 2013 11:49:54 +0100
From: Tamas Nepusz <address@hidden>
To: address@hidden
Subject: Re: [igraph] scale free network

> I need to know is "barabasi.game" able to generate scale free networks with
> special exponent element(e.g. gamma=2.5)? how?

You cannot specify an exact exponent because you will inevitably get a
skewed distribution due to finite-size effects. However, _in_ _theory_, you
could play around with the power=... argument of barabasi.game. Read
Barabasi's paper about nonlinear preferential attachment to understand how
the exponent in the preferential attachment rule influences the exponent of
the degree distribution.

On the other hand, if you want to generate a network whose degree
distribution follows a power-law with gamma=2.5, use static.power.law.game,
where you can specify the exponent explicitly, although in this case the
generation mechanism is different from that of the Barabasi-Albert model.
Finally, you can even generate power-law distributed random integers
according to whichever power-law model you prefer and then use
degree.sequence.game (typically with method="vl") to construct a graph with
exactly the specified degrees.

--
T.