|
From: | Martin Tomko |
Subject: | Re: [igraph] errors using power.law.fit |
Date: | 16 Nov 2010 11:07:51 +0100 |
User-agent: | Mozilla/5.0 (Windows; U; Windows NT 5.1; en-US; rv:1.9.1.8) Gecko/20100227 Lightning/1.0b1 Thunderbird/3.0.3 |
hi Tamas,thanks for your response. How would one call the C code from R? I may be able to install the C library. I would then have to somehow link it to my R installation and send it the graph object or the matrix with degrees that I get as result of degree(g), and get back an indication that I can or cannot fit it with the power.law.fit function, right?
Or would the C method do the Kolmorgov-Smirnov test AND the fitting. thus if I wanted to do it all in R, I just use a standard Kolmorgov-Smirnov?
Also, the article seems to imply - in the recipe part - that you estimate the alpha and Xmin parameters, and then test with the Kolmorgov Smirnov test for goodness of fit. Now, this implies that I still may run into troubles with the optim() error I got our of R. would you have any suggestion how to get rid of that?
Thanks Martin On 11/16/2010 10:07 AM, Tamas Nepusz wrote:
Dear Martin,Now, it is possible that some of the graphs I have would not fit a power law well - is there a recommended way for checking whether it makes sense at all to try fitting one?You can use a Kolmogorov-Smirnov test for that, as outlined here: http://arxiv.org/abs/0706.1062 There is also a plain C implementation of their method which will eventually be integrated with igraph (once I find time for it). Currently it is a separate project, but it works from the command line, so you can simply call out from R to this implementation if you want to do batch processing: https://github.com/ntamas/plfit (Instructions: download, cmake, make, make install).
[Prev in Thread] | Current Thread | [Next in Thread] |