A $d$dimensional copula is a distribution on $[0,1]p^d$ such that its marginals are uniforms on $[0,1]$ A theorem due to Sklar allows to express the law $H$ of a $d$dimensional vector through the action of the unique copula $C$ on its marginals $F_i$ $$ H(x_1,\ldots,x_d) = C(F_1(x_1),\ldots,F_d(X_d)) $$ as soos as those marginals are continuous. We propose here an estimator of this copula $C$ or rather its density $$ c(u_1,\ldots,u_d) = \frac{h(F_1^{1}(u_1),\ldots,F_d^{1}(u_d))}{ f_1(F_1^{1}(u_1))\cdots f_d(F_d^{1}(u_d))}, $$ whose existence is supposed, by a wavelet thresholding method similar to the one used in density estimation.
By default, only communications in english are shown. Using the language filter on the right side, you can also see the french ones.

