Minimal simulation config for a set of paper figures

Folds the selected manifest entries into a scenarios_config covering exactly the case studies, sample size factors and methods they need. Each case study is restricted to its own factors through case_study_sample_size_factors, rather than every case study being crossed with every factor. The fidelity settings do not scale with the selection: ndrift stays at 30 because forest_plot() selects the three principal treatment-effect scenarios by nearest grid point, so a coarser grid would quietly plot different drift values rather than failing.

No paper figure varies the source denominator change factor or the target-to-source standard deviation ratio, so both are pinned to 1.

Usage

paper_replication_requirements(ids, case_studies_config_dir)

Arguments

ids

Manifest ids to cover.

case_studies_config_dir

Directory holding the case study YAMLs.

Value

A scenarios_config list, ready for save_environment().