CopulaModel-package: CopulaModel: Dependence Modeling with Copulas

In YafeiXu/CopulaModel: CopulaModel: Dependence Modeling with Copulas

Description

CopulaModel is the accompanying software for the book: Dependence Modeling with Copulas, by H. Joe, Chapman & Hall/CRC, 2014. With this software, a reader can check (almost) all numerical computations in the book. Much of the contributions for factor copula models is also described in the Ph.D. thesis of P. Krupskii (2014).

Details

All of the algorithms in Dependence Modeling with Copulas are shown within this software. There are templates for doing many numerical calculations and maximum likelihood estimation with copulas. Included are examples of faster code with links to Fortran90 and C. It is not possible to provide functions for all possible uses of copulas. But a user can adapt the templates and code within this software for many other applications. The source code (R, Fortran90, C) has more documentation than these help pages. The functions do not check on the domains of the arguments such as the copula parameter, but the domains are indicated in the source code. If a function doesn't seem to be working with your inputs, check the source code and look for examples in the tests subdirectories and examples in the book chapter subdirectories.

Some features include:

• For the 1-, 2-, 3-parameter bivariate copula families in the book, most have pcop, dcop, rcop, pcondcop, qcondcop for the copula cdf, copula density, copula random generation, conditional distribution C_{2|1} and conditional quantile C_{2|1}^{-1}, where 'cop' is an abbreviated copula name. Also for many bivariate copula families, there are conversions among copula parameter, Kendall's tau, Spearman's rho, Blomqvist's beta, correlation of normal scores, and tail dependence parameters.

• R-vine (regular vine) models for continuous data with specified vine array and parametric pair-copula families.

• R-vine models for discrete response data with possibility of covariates.

• Factor copula models for continuous response (factor models here are truncated vines with latent variables).

• Factor copula models for ordinal (item) response.

Author(s)

H. Joe and P. Krupskii.

References

H. Joe (2014). Dependence Modeling with Copulas. Chapman & Hall/CRC. Boca Raton, FL.

P. Krupskii (2014). Structured Factor Copulas and Tail Inference. PhD thesis, University of British Columbia.

Other specific references are given within code files or help pages.

YafeiXu/CopulaModel documentation built on May 9, 2019, 11:07 p.m.