Calibrate topic proportions based on sufficient statistics — asap_topic_pmf • asapR

Calibrate topic proportions based on sufficient statistics

Usage

asap_topic_pmf(
  beta_dk,
  R_nk,
  Ysum_n,
  a0 = 1,
  b0 = 1,
  max_iter = 10L,
  NUM_THREADS = 0L,
  do_stdize_r = TRUE,
  verbose = TRUE
)

Arguments

beta_dk: dictionary matrix (feature D x factor K)
R_nk: correlation matrix (sample N x factor K)
Ysum_n: sum vector (sample N x 1)
a0: gamma(a0, b0) (default: 1)
b0: gamma(a0, b0) (default: 1)
max_iter: maximum iterations (default: 10)
NUM_THREADS: number of parallel threads (default: 1)
do_stdize_r: standardize correlation matrix R (default: FALSE)
verbose: (default: TRUE)

Value

a list that contains:

beta (D x K) matrix
theta (N x K) matrix
log.theta (N x K) log matrix
log.theta.sd (N x K) standard deviation matrix