Copulas (Q6114814)

Let \(g(x_1,x_2,\dots,x_s)\) be a distribution function defined on \([0,1]^s\) such that every \(k\)-dimensional face function satisfies \[ g(1,\dots,1,x_{i_1},1,\dots,1,x_{i_2},1,\dots,1,x_{i_k},1,\dots,1)=x_{i_1}x_{i_2}\dots x_{i_k}. \] For \(k=1\), these distribution functions are called copulas, which were introduced by \textit{M. Sklar} [Publ. Inst. Statis. Univ. Paris 8, 229--231 (1959)]. For \(k=2\), a distribution function \(g(x,y)\) defined in \([0,1]^2\) is called copula, if \(g(x,1)=x\) and \(g(1,y)=y\) for every \(x, y\). In this paper all known parts of copulas are summarized and extended.