In jakedilliott/covidfireMASS: Agent-Based Model For COVID-19 Dynamics In A Wildfire Season

knitr::opts_chunk$set(
  collapse = TRUE,
  comment = "#>",
  fig.path = "man/figures/README-",
  out.width = "100%"
)

covidfireMASS


Covid Fire Multi Agent Seasonal Simulation

An Agent-Based Model that simulates COVID dynamics within wildfire firefighter camps. This model assumes a hub-and-spoke camp model, which implements module based infection dynamics and quarantine protocols. This work is an extension of the Susceptible Infected Recovered(SIR) model, adding Exposed, Asymptomatic, and Quarantined agent health states. Hence the SEIRAQ model.

Installation

You can install the development version of covidfireMASS from GitHub with:

if (!require("remotes", character.only = TRUE)) {
  install.packages("remotes", dependencies = TRUE)
}
remotes::install_github("jakedilliott/covidfireMASS")

Necessary Data

Incident Assignments

• Each row is a unique fire fighter/resource
• res_id: numeric, resource ID number
• res_gacc: character, resource home GACC (Geographic Area Coordination Center)
• Each column to the right of the second column is labeled with a date and each cell contains an incident that shows which fire each resource was assigned to on that day. On days where a resource was not on a fire inc_id needs to be 0.

Example: Resource 1001 was on incident 1046 from 01/01/2020 - 01/01/2020

library(knitr)
df = data.frame(
  'res_id' = c(1001, 1002, 1003),
  'res_gacc' = c('NM-SWC', 'OR-NWC', 'OR-NWC'),
  '2020-01-01' = c(1046, 1055, 0),
  '2020-01-02' = c(1046, 1055, 0),
  '2020-01-03' = c(1046, 0, 1055),
  '...' = c('...', '...', '...')
)
kable(df, caption = "Incident Assignments Table")

Module Assignments

• res_id and res_gacc columns here are the same as the module assignments data
• Each column to the right of the second column is labeled with a date and each cell contains a module ID that describes which module (team or crew) each resource was assigned to each day. On days where a resource is not on a fire mod_id needs to be 0. These module ID's are only unique when paired with the incident of each agent on that day, there are modules with the same ID across multiple fires.

Example: Resource 1001 was on module E-1 from 01/01/2020 - 01/03/2020

df = data.frame(
  'res_id' = c(1001, 1002, 1003),
  'res_gacc' = c('NM-SWC', 'OR-NWC', 'OR-NWC'),
  '2020-01-01' = c('E-1', 'O-12', '0'),
  '2020-01-02' = c('E-1', 'O-12', '0'),
  '2020-01-03' = c('E-1', '0', 'C-8'),
  '...' = c('...', '...', '...')
)
kable(df, caption = "Module Assignments Table")

Incident Information

| inc_id | inc_number | inc_name | inc_gacc | inc_lat | inc_lon | max_team_type | first_day | last_day |

Minimal Example

library(covidfireMASS)

# Import Data
# This example uses a '.rda' file that includes preconfigured input data
load("sim_inputs_2017_complete.rda")

# Running simulations
# Default simulation run, see ?seasonal_sim for default inputs
sim1 <- seasonal_sim(inc_id_2017, mod_id_2017, inc_info_2017, overhead_ids_2017,
                     vax_df = vax_plan_base_2017)

# seasonal_sim outputs a data frame with the following columns when the 
# `raw` option is TRUE:
#
# res_id     res_gacc     inc_id     mod_id        leader    state
# "numeric"  "character"  "numeric"  "character"  "logical"  "character"
#
# quarantine  q_days     vaccinated  vax_rate   time 
# "logical"   "numeric"  "logical"   "numeric"  "numeric"

# If the `raw` options is set to FALSE (the default), a summarised
# output in a wider format will be returned

jakedilliott/covidfireMASS documentation built on Dec. 20, 2021, 8:59 p.m.