copula.prob: Copula probabilities (joint and conditional) from a fitted...

In GJRM: Generalised Joint Regression Modelling

Copula probabilities (joint and conditional) from a fitted simultaneous model

Description

copula.prob can be used to calculate the joint or conditional copula probabilities from a fitted simultaneous model with intervals obtained via posterior simulation.

Usage

copula.prob(x, y1, y2, y3 = NULL, newdata, joint = TRUE, cond = 0,
            intervals = FALSE, n.sim = 100, prob.lev = 0.05, 
            theta = FALSE, tau = FALSE, min.pr = 1e-323, max.pr = 1)

Arguments

x	A fitted gjrm object as produced by the respective fitting function.
y1	Value of response for first margin.
y2	Value of response for second margin.
y3	Value of response for third margin if a trivariate model is employed.
newdata	A data frame with one row, which must be provided.
joint	If TRUE then the calculation is done using the fitted joint model. If FALSE then the calculation is done from univariate fits.
cond	There are three possible values: 0 (joint probabilities are delivered), 1 (conditional probabilities are delivered and conditioning is with the respect to the first margin), 2 (as before but conditioning is with the respect to the second margin).
intervals	If TRUE then intervals for the probabilities are also produced.
n.sim	Number of simulated coefficient vectors from the posterior distribution of the estimated model parameters. This is used for interval calculations.
prob.lev	Overall probability of the left and right tails of the probabilities' distributions used for interval calculations.
theta	If TRUE the theta dependence parameter will be shown. This is especially useful for prediction purposes when theta is specified as a function of covariate effects.
tau	If TRUE the Kendall's tau will also be calculated and provided in output. Note that the calculation adopted here assumes continuous margins. In all other cases, this may provide a rough indication of dependence under certain assumptions. Note that, for the F, PL and J0 (and the related rotations), computing times may be longer than for the other cases. This is especially useful for prediction purposes when theta is specified as a function of covariate effects, with an interest in analysing a more interpretable measure of dependence for certain copulae.
min.pr, max.pr	Allowed minimum and maximum for estimated probabities.

Details

This function calculates joint or conditional copula probabilities from a fitted simultaneous model or a model assuming independence, with intervals obtained via posterior simulation.

Value

res	It returns several values including: estimated probabilities (p12), with lower and upper interval limits (CIpr) if intervals = TRUE, and p1, p2 and p3 (the marginal probabilities).

Author(s)

Maintainer: Giampiero Marra [e-mail]

See Also

GJRM-package, gjrm

GJRM documentation built on Oct. 25, 2024, 5:07 p.m.