create_projection_object() and Projection_object now take a single migration flag instead of separate emigration and immigration flags. Only emigration was ever read internally, so the two always acted as one switch.
light_migration and light_switching now default to TRUE everywhere, including test_object_generator(). This matches the existing defaults of create_projection_object() and base_object_generator(). Migration and switching output is therefore stored in abbreviated form unless light_migration = FALSE / light_switching = FALSE is set explicitly.
base_object_generator() and create_projection_object() now default to steps = 10 instead of steps = 11, so default projections end in 2070 instead of 2075. Set steps = 11 explicitly to reproduce earlier results.
BUG FIXES
Fixed project() failing with no_diff = TRUE because ASFR output lacked religion column names.
The Total column of the ASFR output now reports the realised rate (births / exposure) instead of the input ASFR. Under combined fertility factorization the two differ, so Total previously did not match the saved births.
Realised ASFRs (Total and all religion columns of the "total" structure) are now 0 where exposure is 0, instead of the unweighted mean of the structure rates. *_post_switch columns keep the mean of the applied rates, and religion columns of individual structures still report the applied rate. ASFR output again matches the May 2026 projection exactly.
Fixed R CMD check “no visible binding for global variable” notes: .data is now imported from ggplot2, plotting code uses .data$ for column names in aes(), and fold_asfr uses data.table::set() instead of := for Mother_age_group.
DOCUMENTATION
Documented the ASFR output convention for previous_asfr: Total is the realised rate (births / exposure), religion columns of individual structures hold the applied rates, and realised rates are 0 where exposure is 0.
Documented that combined_fertility_factorization defaults to TRUE (field docs and the inner-workings vignette).
Updated docs, examples and vignettes for the single migration flag, the new light_migration / light_switching defaults and the new default number of projection steps.
Examples now use named arguments (e.g. project(projection_object = obj), test_object_generator(object_name = ...)).
Dataset docs: Sex is described as numeric (1 = males, 2 = females) and data_population as 201 rows.
Corrected the flat_fertility description, the outmigration_rates column description, the Projection_object class description and the region field descriptions (with examples).
Updated the package description (CASPAR, Pew-Templeton Global Religious Futures project) and spelled “Pew” consistently.
README: the introduction now matches the package description, and the examples were rewritten to show a full projection (step by step and with base_object_generator()), exporting results, and a quick run with test_object_generator().
measuringreligiouschange 1.0.1 (2026-11-09)
BUG FIXES
Enabled combined fertility factorization by default
Changed data used in testing to match Pew report
Fixed issue causing age group 0 to have incorrect mortality saved