Abstract
We provide a method to construct a class of n-copulas C defined as C(u) = D(u_1 , . . . , u_n-1 ) u_n + A(u_1 , . . . , u_n-1) f(u_n). These copulas are obtained from a (n-1)-copula D and some suitable auxiliary functions A and f. Members of this class have been fully characterized in terms of properties of the auxiliary functions. The proposed copulas may model some weak dependence and a non-exchangeable behavior among the components of a random vector. Moreover, they can be easily simulated and fitted to real data.