consider the case where we have a survey sampling design with $S$ strata indexed by $s \in 1:S$, each with $C_s$ clusters indexed by $c \in 1:C_s$.

Assume that we are given as input a data.frame containing the survey sampling probability for each cluster in each stratum, and assume that each serological observation has a cluster ID and stratum ID.

Add an option in est_seroincidence() to adjust the seroincidence likelihood calculations for this survey sampling design.