Fig. 7.

DLBCL Data; posterior distributions $\text{Pr}(K_{+}|\mathbf{y})$ of the number of clusters $K_{+}$ for mixtures of skew normal (left-hand panel) as well as mixtures of skew-t distributions (right-hand panel); top row in each panel: sparse finite mixtures with $K=20$, $e_0\sim \mathcal{G}(1,100)$ (left column) and matched prior $e_0\sim \mathcal{G}(2,4K)$ (right column); bottom row in each panel: DPM with $\alpha \sim \mathcal{G}(2,4)$ (right column) and matched prior $\alpha \sim \mathcal{G}(1,100/K)$ (left column)