Closed cschwarz-stat-sfu-ca closed 3 years ago
JAGS appears to handle n1=0 properly everywhere in the code. Related to bad n1, bad m2 strata where n1 and m2 are forced to be 0. Then m2=0 recoveries out of n1=0 releases is binomial or multinomial and JAGS does the computations properly.
If n1=0, we currently set n1 to 1 because OpenBugs didn't like binomials with n=0 and x=0. Perhaps this is not resolved with JAGS.