When most people talk about Power Law they always allude to the Pareto distribution which underestimates the inequality of lower quantiles when compared to the upper quantile, but that most distributions (e.g. YouTube, citation networks, blogs) more or less follows the log-normal or exponential distribution which closely matches this change.
- Why does this mismatch of power law vs false power law (or broken power law) still exist?
- What can cause such a discrepancy regarding scale free vs modular networks?
https://aaronclauset.github.io/powerlaws/data.htm
When most people talk about Power Law they always allude to the Pareto distribution which underestimates the inequality of lower quantiles when compared to the upper quantile, but that most distributions (e.g. YouTube, citation networks, blogs) more or less follows the log-normal or exponential distribution which closely matches this change.
https://aaronclauset.github.io/powerlaws/data.htm