Changes in version 0.2.0 (2026-10-07)                  

  - The integer programs are solved with HiGHS (the highs package)
    instead of lpSolve, which returned infeasible points as optimal for
    some inputs. Reach and frequency are fixed with exact bounds, and
    the penetration stage is decided in R, so the tie-breaks are exact
    up to double-precision rounding.
  - After its first penetration solve, turf() searches for other
    portfolios with the same penetration, within the solver tolerance.
    That search stops after 1,000 more portfolios or 30 seconds. If it
    stops early, turf() gives a warning, and the penetration is optimal
    only to within the solver tolerance (about 1e-9).
  - turf_min_cover() checks the solver's result and stops with an error
    if the selection does not cover everyone or is not integral.
  - turf_simulate() checks its arguments.
  - The repository also contains Python (python/) and JavaScript (js/)
    ports that follow the same specification (docs/algorithm.md) and
    pass the same conformance suite (conformance/), and a one-file
    browser dashboard (dashboard/) that runs the JavaScript port. They
    are not part of the R package.

                        Changes in version 0.1.0                        

First release. The package replaces the 2021 prototype script.

  - turf() finds the portfolio of a given size with maximum reach, with
    frequency and penetration tie-breaks.
  - turf_min_cover() finds the smallest portfolio that reaches every
    reachable respondent.
  - turf_sizes() solves a range of portfolio sizes and returns a table.
  - turf_simulate() makes random reach data for examples and tests.
  - Example data from public studies: ham, coffee, pies, and lunchbags.
    Simulated example data of three sizes: icecream, chips, and cafe.