copulaSample: Generate correlated samples from elicited marginal...

In SHELF: Tools to Support the Sheffield Elicitation Framework

Generate correlated samples from elicited marginal distributions using a multivariate normal copula

Description

Takes elicited marginal distributions and elicited concordance probabilities: pairwise probabilities of two uncertain quantities being greater than their medians, and generates a correlated sample, assuming the elicited marginal distributions and a multivariate normal copula. A vignette explaining this method is available at https://oakleyj.github.io/SHELF/Multivariate-normal-copula.html

Usage

copulaSample(..., cp, n, d = NULL, ex = 1)

Arguments

...	A list of objects of class elicitation. command, one per marginal distribution, separated by commas.
cp	A matrix of pairwise concordance probabilities, with element i,j the elicited probability P(X_i > m_i, X_j > m_j or X_i < m_i, X_j < m_j), where m_i and m_j are the elicited medians of the uncertain quantities X_i and X_j. Only the upper triangular elements in the matrix need to be specified; the remaining elements can be set at 0.
n	The sample size to be generated
d	A vector of distributions to be used for each elicited quantity: a string with elements chosen from "normal", "t", "gamma", "lognormal", "logt", "beta", "mirrorgamma", "mirrorlognormal", "mirrorlogt". The default is to use the best fitting distribution in each case.
ex	If separate judgements have been elicited from multiple experts and stored in the elicitation objects, use this argument to select a single expert's judgements for sampling. Note that this function will not simultaneously generate samples for all experts.

Value

A matrix of sampled values, one row per sample.

Author(s)

Jeremy Oakley [e-mail]

Examples

## Not run: 
p1 <- c(0.25, 0.5, 0.75)
v1 <- c(0.5, 0.55, 0.6)
v2 <- c(0.22, 0.3, 0.35)
v3 <- c(0.11, 0.15, 0.2)
myfit1 <- fitdist(v1, p1, 0, 1)
myfit2 <- fitdist(v2, p1, 0, 1)
myfit3 <- fitdist(v3, p1, 0, 1)
quad.probs <- matrix(0, 3, 3)
quad.probs[1, 2] <- 0.4
quad.probs[1, 3] <- 0.4
quad.probs[2, 3] <- 0.3
copulaSample(myfit1, myfit2, myfit3, cp=quad.probs, n=100, d=NULL)

## End(Not run)

SHELF documentation built on Sept. 11, 2024, 6:54 p.m.