Statalist


[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

st: RE: Estimating Power law distributions


From   "Nick Cox" <n.j.cox@durham.ac.uk>
To   <statalist@hsphsun2.harvard.edu>
Subject   st: RE: Estimating Power law distributions
Date   Mon, 27 Apr 2009 11:31:05 +0100

You have at least three main choices: 

1. Translate this into a probability density function and estimate using
your method of choice, say maximum likelihood. 

2. Use -cumul- to calculate the distribution function and thus the
complement explicit here. Then estimate the parameters e.g. by least
squares after transformation. 

3. Use graphs. See for example detailed comments on power laws within 

SJ-5-3  gr0018  . . . . . . . . . .  Speaking Stata: The protean
quantile plot
        . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .  N.
J. Cox
        Q3/05   SJ 5(3):442--460           (see gr41_3 and gr42_3 for
commands)
        discusses quantile and distribution plots as used in
        the analysis of species abundance data in ecology

Incidentally, none of these methods requires any kind of normalisation. 

Nick 
n.j.cox@durham.ac.uk 

Emanuele Canegrati, Ph.D.

I have to check if the cumulative distribution of a series, say Sj, is
consistent with a power law, say Prob{S>x} distas x*exp(-aSj). I should
normalize my series by transforming it to zero mean and unit s.d. Then I
should evaluate the value for the exponent aSj. Do you know how to do
it? 
 

*
*   For searches and help try:
*   http://www.stata.com/help.cgi?search
*   http://www.stata.com/support/statalist/faq
*   http://www.ats.ucla.edu/stat/stata/



© Copyright 1996–2014 StataCorp LP   |   Terms of use   |   Privacy   |   Contact us   |   What's new   |   Site index