Services & Resources / Wolfram Forums / MathGroup Archive
-----

MathGroup Archive 2011

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

Search the Archive

References for ZipfDistribution and References in help pages in general

  • To: mathgroup at smc.vnet.net
  • Subject: [mg122794] References for ZipfDistribution and References in help pages in general
  • From: ham sha <hamsha.ir at gmail.com>
  • Date: Thu, 10 Nov 2011 06:56:21 -0500 (EST)
  • Delivered-to: l-mathgroup@mail-archive0.wolfram.com

Hello everyone,

I am going to use ZipfDistribution[Rho] function for doing some
statistical calculations but its implementation has confused me and
there is no references for it.

In its definition they write:

The probability for value x and finite n is given by x^-(\[Rho]+1)/
Subsuperscript[H, n, (\[Rho]+1)] and for infinite n given by x^-(\[Rho]
+1)/\[Zeta](\[Rho]+1).

The definition of the infinite case is different from what one usually
finds in technical papers and the internet, e.g. at wikipedia
http://en.wikipedia.org/wiki/Zipf%27s_law .

For infinite case, the exponent is 1+Rho in the Mathematica definition
while for other papers there is no 1 .

Unfortunately Wolfram team hasn't provided any references for its
implementation and I am confused by it.

Does someone has any suggestions regarding this problem, or does
someone knows a reference to clarify the inconsistency of definitions?

Generally this is a big problem by the help system of the Mathematica.
For another system one often finds a Reference section at the end of help
pages which  allows users to see the definition and the algorithms
used in the implementation. This is unfortunately not true with the
Mathematica which is very a powerful system.

Thank you in advance,
Sincerely,
Hamed



  • Prev by Date: Question on two-stage optimization of polynomials: how to draw a response function?
  • Next by Date: Re: FinancialData - how to find ticker symbols?
  • Previous by thread: Re: Question on two-stage optimization of polynomials:
  • Next by thread: More Efficient Use of Switch Statement