A practical recipe to fit discrete power-law distributions

摘要: Power laws pervade statistical physics and complex systems, but, traditionally, researchers in these fields have paid little attention to properly fit distributions. Who has not seen (or even shown) a log-log plot of completely curved line pretending be power law? Recently, Clauset et al. proposed method decide if set values variable distribution whose tail is law. The key their procedure the identification minimum value for which holds, selected as Kolmogorov-Smirnov distance between empirical its maximum-likelihood minimum. However, it been shown that this can reject power-law hypothesis case simulated data. Here we propose simpler selection criterion, illustrated with more involving discrete