The evidence lower bound
elbo(
warray,
lbeta_array,
cor_inv,
postmean,
postvar,
bias,
seq,
mean_bias,
var_bias,
mean_seq,
var_seq,
ploidy
)An three-way array. The (i,j,k)th entry is the variational posterior probability
that individual i at SNP j has dosage k - 1. See compute_all_post_prob.
A three-way array. The (i,j,k)th entry is the log-density of the betabinomial
for individual i at SNP j and dosage k - 1. See compute_all_log_bb.
The inverse of the correlation matrix.
A matrix. The (i,j)th entry is the variational posterior mean for individual i at SNP j.
A matrix. The (i,j)th entry is the variational posterior variance for individual i at SNP j.
A vector. The jth entry is the allele bias for SNP j.
A vector. The jth entry is the sequencing error rate at SNP j.
The prior mean on the log-bias.
The prior variance on the log-bias.
The prior mean on the logit of the sequencing error rate.
The prior variance on the logit of the sequencing error rate.
The ploidy of the species.
A double. The evidence lower-bound that mupdog
maximizes.