Pairing properties of cross-ranked multivariate incidence and intensity distributions

Autor:	Ganapati P. Patil, Wayne L. Myers
Rok vydání:	2015
Předmět:	Statistics and Probability Multivariate statistics Multivariate analysis Ranking Rank (linear algebra) Joint probability distribution Pairing Statistics Statistics Probability and Uncertainty Cluster analysis Data matrix (multivariate statistics) General Environmental Science Mathematics
Zdroj:	Environmental and Ecological Statistics. 22:651-665
ISSN:	1573-3009 1352-8505
Popis:	Sequentially ordered placement of cross-ranks is used to explore evidence of strong structure in multivariate distributions of incidence and/or intensity variates. Cross-ranking consists of ranking the ranks in observational rows after first ranking the values in column variates. Strong structure is viewed as being expressed in conditionally conjunctive characteristics of joint distribution for variates. Such structure can be internal to the data matrix and/or external as arrangement in space or on a network. In exploration of a study situation involving ecological conservation characteristics, the pairing properties of placements for cross-ranks reveal strong structure that is both internal to the data matrix and external as geographic clustering but not immediately evident from conventional rank correlations. The R statistical software system is used for analysis, and special R functions are included.
Databáze:	OpenAIRE
Externí odkaz:	https://explore.openaire.eu/search/publication?articleId=doi_________::7181b434e3fe7c5f3260fff9f8137e46 https://doi.org/10.1007/s10651-015-0324-9 Zobrazit plný text záznamu Full text from SpringerLink