loglike_cond_fast: Calculate the log likelihood of a cluster conditioned on size...

In skgallagher/TBornotTB: Analyze Tuberculosis Patients in Maryland

Description

Calculate the log likelihood of a cluster conditioned on size and smear pos. num

Usage

loglike_cond_fast(p_plus, p_neg, clust_df, is_truncated = FALSE, w_pos = 0.5)

Arguments

p_plus	probability of smear pos
p_neg	probability of smear neg
clust_df	data frame with the following columns
is_truncated	logical default is FALSE
w_pos	empirical weights of number of positive smears i_pos number of transmission from a smear positive person i_neg number of transmission from a smear negative person (optional) n_pos number of positive smears n cluster size freq frequency of time this particular cluster occurs

Details

if i_neg is not a column in clust_df we will use n-i_pos - 1 as i_neg. If 'is_truncated' is TRUE, we will calculate the the zero-truncated likelihood

Value

log likelihood summed over the clusters

skgallagher/TBornotTB documentation built on April 21, 2020, 1:19 p.m.