epidemics 0.3.0

This is a minor version release of epidemics is the end point of this project that allows vector inputs to the Ebola model (#211, #212).

Breaking changes

1. The Ebola model run with a single parameter set and a single set of composable (i.e., functionality before this vignette) now runs 100 replicates of the model by default, and the output returns the additional column name "replicate" (#211).

2. The Ebola model disallows rate interventions on, and time-dependence for, the 'infectiousness rate' and the 'removal rate' as these control the number of epidemiological sub-compartments and cannot be allowed to vary during run time.

3. The default behaviour of epidemic_size() is to exclude the 'dead' compartment from epidemic size calculations; this has changed from including it by default, as most models don't have a 'dead' compartment (#212);

Model functions

1. The Ebola model code has been split into a two-level structure similar to the ODE models. The user facing function model_ebola() now handles input checking and some cross-checking, and makes combinations of infection parameter sets and scenarios, and handles output. The internal model function .model_ebola_internal() is now called over combinations of parameters and scenario elements (#211).

2. .model_ebola_internal() relies on functionality from the {withr} package to preserve the random number seed between parameter set-scenario combinations, and to ensure that each $i$-th replicate of each scenario starts with the same random number stream.

Model structures

Classes

1. The <intervention> helper functions .cumulative_rate_intervention() and .intervention_on_rates() have been renamed to include the prefix . to indicate they are internal functions (#211).

2. Small updates to .cross_check_intervention() to better handle intervention sets where contacts interventions are optional (the Ebola model) (#211).

Helper functions

1. .check_prepare_args_ebola() is a new argument cross-checking and preparation function for the Ebola model (#212).

2. epidemic_size() is substantially updated (#212):

   • Added option for time which returns epidemic size at a specific time point, overriding the stage argument, defaults to NULL as the intended use of the function is to return the final size;

   • Added option to return epidemic sizes at multiple stages or time points (stage and time can be vectors);

   • Added option to simplify the output to a vector, which is TRUE by default to keep consistency with previous functionality;

   • Added functionality to handle replicates from the Ebola model;

   • Added tests for new functionality.

3. Added function .output_to_df_ebola() to handle output from the Ebola model (#211).

Documentation

1. Updates to the Ebola model vignette showing new functionality.

2. Updates to the design decisions vignette documenting decisions taken for the Ebola model.

3. Updated sections on handling social contacts data in the 'Getting started' and 'Modelling interventions on social contacts' vignettes, added reference to Wallinga et al. 2006 https://doi.org/10.1093/aje/kwj317 (#217).

4. Enables auto development mode for the website in _pkgdown.yml, and manually specify Bootstrap version 5 (#213).

5. Updates WORDLIST.

6. Updates installation instructions in the Readme; link under 'Contribute' correctly directs to pull requests page (#217).

Package

1. {withr} moved from Suggests to Imports due to use in seed management.

2. Added @bahadzie as contributor and @jamesmbaazam as reviewer.

3. Updates to the GitHub Actions workflows and linter config file.

4. Updates the license year to 2024 in all files with the year.