Skip to contents

malariasimple 0.1.0

New features

  • Added a stochastic version of the model, selectable via get_parameters(stochastic = TRUE).
  • Added the ability to propagate parameter uncertainty by drawing from the joint posterior distribution via the parameter_draws argument of get_parameters() (defaults to "median", which reproduces the previous central estimates). See the new parameter_draws_df dataset.
  • Added a fully correlated option for ITN distribution via set_bednets(distribution_type = "correlated") (default remains "random").
  • Added seasonal malaria chemoprevention (SMC) via set_smc().
  • Added support for a time-varying extrinsic incubation period (eip).
  • Added custom output post-processing helpers (get_custom_output(), make_2d()).
  • run_simulation() now errors clearly if set_equilibrium() has not been called.

Bug fixes

  • Corrected the ITN insecticide decay rate to use the mean retention time (rate 1 / gamman), matching malariasimulation’s bednet_decay().
  • Fixed a double-counting error in ITN decay when only a single distribution event was specified.

Breaking changes

These changes alter model behaviour and/or the public interface. Existing scripts will need updating, and numerical results may differ from previous versions.

  • Several get_parameters() arguments were renamed:
    • het_brackets -> biting_groups
    • ft -> daily_ft
    • delayMos -> eip
    • mu0 -> mum
    • tau1 / tau2 -> foraging_time / gonotrophic_cycle
    • bites_Bed / bites_Indoors -> phi_bednets / phi_indoors
    Old argument names are no longer recognised and are silently ignored.
  • Some default parameter values changed: tsd (4 -> 3), number of biting heterogeneity groups (5 -> 4), phi_bednets (0.89 -> 0.85), phi_indoors (0.97 -> 0.9). A new human_pop argument (default 100000) now scales the initial human state sizes.
  • set_seasonality() now takes Fourier coefficients g0, g and h directly instead of a season_params object. The get_seasonal_forcing() and get_peak_cc() signatures changed accordingly.