You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Improved prevalence computation code, it now consumes slightly less time and is better organized. In addition test-coverage and documentation is improved.
Cluster/strata/sw values in anthro_prevalence now support an explicit NULL value to better control for (missing) cluster/strata/sw information.
New API functions are exported, mainly useful to use complicated portions of the logic in other packages, not meant for end-users.
nest = TRUE is now set for the survey designs in anthro_prevalence. See the documentation of survey::svydesign for more information.
Removed headc, armc, triskin and subskin from parameters of anthro_prevalence as they were not used for the final output.
_pop/unwpop values in prevalence calculation are now 0 instead of NA if no values are present for that group.
Bugfix
Previously, if one level in a group only has one observation, all stdev columns in anthro_prevalence are NA. Now only the levels with one observations yield NA, for the others the stdev is still computed.