The goal of empCop is to give tools to work with Empirical checkerboard copula, empirical checkerboard copula with known margins and convex combinaisons of copulas. It uses the main classes and formal generics from the copula package. It aims at implmeenting a lot of empirical model to asses multivaraite dependance structures that are not already implement in the main copula package.

## Installation

WARNING : This package is in active developpement an is clearly NOT stable. use with caution. If you are still not scared, you can install the released version of empCop from Github with :

devtools::install_github("lrnv/empCop")

## Example

Starting with the LifeCycleSavings dataset, we could first plot and then compute the checkerboard copula ith parameter $m=5$ of this dataset with the following code :

data("LifeCycleSavings")
pseudo_data <- (apply(LifeCycleSavings,2,rank,ties.method="max")/(nrow(LifeCycleSavings)+1))
pairs(pseudo_data) # plot pairs of pseudo_data

cop <- cbCopula(x = pseudo_data,m = 5,pseudo = TRUE) # add pseudo=TRUE if you provided pseudo observation

And then this copula can be easily worked with, using for exemple rCopula to simulate, pCopula to calculate it’s values, and other common generics from the copula package.

On the other hand, the empirical checkerboard copula with known margins can be constructed with the following code :

  true_copula <- onacopulaL(
family = "Clayton",
nacList = list(iTau(getAcop("Clayton"), 0.6), 1:4)
)
dataset <- rCopula(50,true_copula)
colnames(dataset) <- c("u","v","w","x")
cop <- cbkmCopula(x = dataset,m = 5,pseudo = TRUE,margins_numbers = c(2,3),known_cop = known_clayton)
And then, as before, this copula can be easily handled in the classical framework of copula’s S4 classes.