Bivariate box plots based on quantile regression curves

Navarro Jorge ()
Additional contact information
Navarro Jorge: Facultad de Matemáticas, Universidad de Murcia, 30100Murcia, Spain

Dependence Modeling, 2020, vol. 8, issue 1, 132-156

Abstract: In this paper, we propose a procedure to build bivariate box plots (BBP). We first obtain the theoretical BBP for a random vector (X, Y). They are based on the univariate box plot of X and the conditional quantile curves of Y|X. They can be computed from the copula of (X, Y) and the marginal distributions. The main advantage of these BBP is that the coverage probabilities of the regions are distribution-free. So they can be selected by the users with the desired probabilities and they can be used to perform fit tests. Three reasonable options are proposed. They are illustrated with two examples from a normal model and an exponential model with a Clayton copula. Moreover, several methods to estimate the theoretical BBP are discussed. The main ones are based on linear and non-linear quantile regression. The others are based on empirical estimators and parametric and non-parametric (kernel) copula estimations. All of them can be used to get empirical BBP. Some extensions for the multivariate case are proposed as well.

Keywords: Median regression; quantile confidence bands; copula; kernel estimation; 62G99; 62G07 (search for similar items in EconPapers)
Date: 2020
References: View references in EconPapers View complete reference list from CitEc
Citations:

Downloads: (external link)
https://doi.org/10.1515/demo-2020-0008 (text/html)

Related works:
This item may be available elsewhere in EconPapers: Search for items with the same title.

Export reference: BibTeX RIS (EndNote, ProCite, RefMan) HTML/Text

Persistent link: https://EconPapers.repec.org/RePEc:vrs:demode:v:8:y:2020:i:1:p:132-156:n:11

DOI: 10.1515/demo-2020-0008

Access Statistics for this article

Dependence Modeling is currently edited by Giovanni Puccetti

More articles in Dependence Modeling from De Gruyter
Bibliographic data for series maintained by Peter Golla ().