Drift ranges and target sample sizes for each case study (table S1)
Source:R/table_case_study_summary.R
table_drift_ranges_and_sample_sizes.RdThe paper's table S1, which merges what used to be two
tables: the drift and treatment-effect ranges, and the target-study
sample sizes. The ranges come from compute_drift_range(), the function
the simulation builds its drift grid with, so the table describes the
grid the current configs produce rather than whatever an older results
directory happens to hold.
The sample-size columns are the nominal totals floor(N_S / k), as the
paper prints them. The simulation puts floor(N_S / (2k)) patients in
each arm, so the total it actually simulates can be one or two lower -
see paper_sample_size_per_arm().
Every case study is listed whatever the selection, because the paper's table always has all six rows.