I wonder if openturns supports customizing a copula, just like customizing a probability distribution.
For example, a copula like C(u,v;\theta) = uv + \theta [u(u-1) (2u-1)][v(v-1)(2v-1)].
This copula is a useful example to show that two random variables are not quadrant dependent. I believe he deserves some inquiry.
Hello,
You can implement your own copula like a distribution,
one notable difference is that isCopula should always return True, see
https://openturns.github.io/openturns/latest/auto_probabilistic_modeling/distributions/plot_python_distribution.html?highlight=create+customized+distribution+copula