Fast four category point estimates

[athowes]

Segmentation fault (core dumped) error now for TZA on aaa_fit_3-multi-sexbehav even though it has run before in the past e.g. report 20211220-120015-7fcc43b3 in the archive :angry:

[athowes]

Need to add 15-24 aggregates for FSW logit model.

[athowes]

Actually, shouldn't be producing aggregates until right at the end.

[athowes]

There are now complete estimates! Details:

• Three category model with IID age, Besag spatial and IID survey effects (when there are multiple surveys) fitted to all DHS and AIS data
• FSW logit model with IID age, Besag spatial, CFSW ever covariate fitted to all surveys with specific FSW question (including PHIA).
• Point estimate from sex non-regular category split into sex non-regular and FSW using FSW logit model.
• FSW category adjusted to match national estimates from Johnston et al. (changing the sex non-regular category to sum to one).