Beta rank function: A smooth double-Pareto-like distribution


Por: Fontanelli, Oscar, Miramontes, Pedro, Mansilla, Ricardo, Cocho, Germinal, Li, Wentian

Publicada: 3 jun 2022 Ahead of Print: 1 ene 2020
Categoría: Statistics and probability

Resumen:
The beta rank function (BRF), x (u) = A(1 - u)(b)/u(a), where u is the normalized and continuous rank of an observation x, has wide applications in fitting real-world data. The underlying probability density function (pdf) f(X)(x) is not expressible in terms of elementary functions except for specific parameter values. We show however that it is approximately a unimodal skewed two-sided power law, or double-Pareto, or log-Laplacian distribution. Analysis of the pdf is simplified when the independent variable is log-transformed; the pdf f(Z = log X)(z) is smooth at the peak; probability is partitioned by the peak with proportion root b/root a (left to right); decay on left and right tails is approximately exponential, e(z/log (A)/b) / b and e(-z-log(A)/a) / a respectively. On the other hand, f(X) (x) behaves like a power distribution x(1/b-1) when x similar to 0 and decays like a Pareto 1/x(1/a+1) when x >> 0. We give closed-form expressions of both pdf's in terms of Fox-H functions and propose numerical algorithms to approximate them. We suggest a way to elucidate if a data set follows a one-sided power law, a lognormal, a two-sided power law or a BRF. Finally, we illustrate the usefulness of these distributions in data analysis through a few examples.

Filiaciones:
Fontanelli, Oscar:
 Centro de Investigaciones Interdisciplinarias en Ciencias y Humanidades, Universidad Nacional Autónoma de Mexico, Mexico City, Mexico

Miramontes, Pedro:
 Departamento de Matemáticas, Facultad de Ciencias, Universidad Nacional Autónoma de Mexico, Mexico City, Mexico

Mansilla, Ricardo:
 Centro de Investigaciones Interdisciplinarias en Ciencias y Humanidades, Universidad Nacional Autónoma de Mexico, Mexico City, Mexico

Cocho, Germinal:
 Instituto de Fisica, Universidad Nacional Autonoma de Mexico, Mexico City, Mexico

Li, Wentian:
 The Robert S. Boas Center for Genomics and Human Genetics, The Feinstein Institutes for Medical Research, Manhasset, NY, United States
ISSN: 03610926





COMMUNICATIONS IN STATISTICS-THEORY AND METHODS
Editorial
TAYLOR & FRANCIS INC, 530 WALNUT STREET, STE 850, PHILADELPHIA, PA 19106 USA, Estados Unidos America
Tipo de documento: Article
Volumen: 51 Número: 11
Páginas: 3645-3668
WOS Id: 000555193600001
imagen Green Submitted

MÉTRICAS