Online appendices for Gray, et al. Hahahahaha, Duuuuude, Yeeessss!: A two-parameter characterization of stretchable words and the dynamics of mistypings and misspellings. (pdf) (arxiv)

Online Appendix B

Token Count Distributions

Select a kernel to display the token count distribution for that kernel. The horizontal axis represents the length (number of characters) of the token and the vertical axis gives the total number of tokens of a given length that match this kernel. The included statistics give the kernel rank, r, the value of the balance parameter (normalized entropy, H), and the value of the stretch parameter (Gini coefficient, G) for this kernel. See Sec. IIIA in the paper for more information.