For following model structure: Create an object of type "HDP2", which represents the model structure: G_m |eta ~ DP(eta,U), m = 1:M G_mj|gamma ~ DP(gamma,G_m), j = 1:J_m pi_mj|G_mj,alpha ~ DP(alpha,G_mj) z|pi_mj ~ Categorical(pi_mj) k|z,G_mj ~ Categorical(G_mj), if z is a sample from the base measure G_mj u|k,G_m ~ Categorical(G_m), if k is a sample from the base measure G_m theta_u|psi ~ H0(psi) x|theta_u,u ~ F(theta_u) where DP(eta,U) is a Dirichlet Process on positive integers, eta is the "concentration parameter", U is the "base measure" of this Dirichlet process, U is an uniform distribution on all positive integers. DP(gamma,G_m) is a Dirichlet Process on integers with concentration parameter gamma and base measure G_m. DP(alpha,G_mj) is a Dirichlet Process on integers with concentration parameter alpha and base measure G_mj. The choice of F() and H0() can be arbitrary, they are distributions of x and theta_u correspondingly. In the case of HDP2, u, z and k can only be positive integers. The sufficient statistics of a set of samples x in a "HDP2" object is the same sufficient statistics of the "BasicBayesian" inside the "HDP2", see examples.
# S3 method for HDP2
sufficientStatistics_Weighted(obj, x, w, ...)A "HDP2" object.
Random samples of the "BasicBayesian" object.
numeric, sample weights.
Additional arguments to be passed to other inherited types.
Return the sufficient statistics of the corresponding BasicBayesian type, see examples.
Teh, Yee W., et al. "Sharing clusters among related groups: Hierarchical Dirichlet processes." Advances in neural information processing systems. 2005.