[](https://codecov.io/gh/grattan/covid19.model.sa2?branch=master)

covid19.model.sa2

A model of the COVID-19 pandemic stratified by SA2.

Synopsis

Basic usage:

S <- simulate_sa2(days_to_simulate = 5)

The default return value is a list of three components:

• integer(days_to_simulate) The number of people infected on each day simulated
• A data.table of 21,364,885 rows, one for each individual modelled, with columns V1, ... Vd where d = days_to_simulate and each column is the status of each individual on each day simulated.
• A vector of test results.

(Other return types are available via returner but these are subject to change.)

Modelling different epidemiological assumptions

To model different epidemiological assumptions -- such as the the duration of the incubation period, how likely transmission is in certain places, and
the severity of cases -- supply a list of parameters to EpiPars.

The function set_epipars() returns a list of the required parameters with reasonable defaults.

S <- simulate_sa2(EpiPars = set_epipars())

# Assume longer average incubation, 8 days
S_long_incubation <- 
  simulate_sa2(EpiPars = set_epipars(incubation_mean = 8))

# Assume everyone in the household gets infected the following day
# if any member does
S_high_household_transmission <- 
  simulate_sa2(EpiPars = set_epipars(q_household = 1))

# Assume no-one is naturally resistant
S_no_natural_resist <-
  simulate_sa2(EpiPars = set_epipars(resistance_threshold = 1000))

Policy parameters

Like EpiPars use set_policypars to supply a list of PolicyPars, to change the assumptions about policies that restrict interaction.

# Open all schools
S_schools <- 
  simulate_sa2(PolicyPars = set_policypars(schools_open = TRUE))
  
## Isolate everyone over 65
S_ages_lockdown <-
  simulate_sa2(PolicyPars = set_policypars(age_based_lockdown = 65:100))

Performance

Two arguments are available to improve the performance of the model, as well as corresponding options.

• nThread the number of threads to use during the modelling
• use_dataEnv can be set to TRUE to avoid boilerplate reading in and preparation of the base data.

Installation

Mac

• Install R 4.0.0: https://cran.r-project.org/bin/macosx/
• Enable OpenMP by installing rtools: https://github.com/rmacoslib/r-macos-rtools/releases/tag/v3.2.2