Simulating multivariate extreme value distributions of logistic type (Q1887258)

A multivariate extreme value distribution is considered with CDF \[ G(x)=\exp \left( -\sum_{b\in B} \left( \sum_{i\in b} \left( {\vartheta_{i,b}\over x_i} \right)^{1/\alpha_b} \right)^{\alpha_b} \right),\quad x_i >0, \] where \(B\subseteq 2^{\{1,\dots,d\}}\), \(\alpha_b\in(0,1)\), \(\vartheta_{i,b}\in[0,1]\) are parameters. \(G\) is called multivariate asymmetric logistic model. The symetric model corresponds to \(B=\{1\}\), \(\vartheta_{i,b}=1\). Different algorithms for random number generation from \(G\) are considered.