A fresh look at Prelims’ degree distribution…is it really scale free?

As Statistical Methods pointed out in their comment on my post, the methodology I used when I proposed that the Prelims graph’s degree distribution was scale free is outdated and not conclusive. This afternoon I decided to take a fresh look at the data following the methodology of Clausset et. al using the Python Powerlaw library. After fitting the data and plotting the probability density function, I evaluated the goodness of fit of the power law distribution through comparisons to the fit of other distributions. The results indicated that a power law distribution may not be the best fit (although a better than an exponential distribution), and a better fit might be a stretched exponential distribution (p > .05). In the following figure you can see the actual data (blue line), a power law fit (red dotted line), a log normal fit (green dotted line), and a stretched exponential fit (blue dotted line). More about this in a couple of weeks.

loglikelihood

Leave a Comment

Filed under Uncategorized

Leave a Reply

Your email address will not be published. Required fields are marked *

You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <s> <strike> <strong>