GitHub

EpiEconShocks.jl

Project Status: Concept – Minimal or no implementation has been done yet, or the repository is only intended to be a limited example, demo, or proof-of-concept. Version Dev Build Status Coverage SciML Code Style

EpiEconShocks.jl is a proof-of-concept package that demonstrates how to translate epidemic outcomes such as cases, hospitalisations, and deaths into 'shocks' or disruptions to the economic system of the affected population.

EpiEconShocks.jl is developed at the Jameel Institute at Imperial College London as part of the Jameel Institute-Kenneth C. Griffin Initiative for the Economics of Pandemic Preparedness (EPPI), in collaboration with the RESIDE research software engineering team.

EpiEconShocks.jl is currently a thin wrapper around the GlobalTradeAnalysisProjectModelV7.jl package which implements the GTAP trade model, with wrappers around other economic models under consideration.

Installation

EpiEconShocks.jl can be installed from GitHub using the Julia package manager Pkg.jl.

using Pkg
Pkg.add(url="git@github.com:jameel-institute/EpiEconShocks.jl.git")

Quick start

Here is a basic example of the current functionality using example GTAP data. A more detailed example may be provided once more detailed data can be shared online.

First, create a ParameterShock object that specifies the parameter set, and the sectors within the set, that are to be shocked. Shocks should be represented as scaling factors in the range $[0.0, 1.0]$; users must take care to convert an estimated reduction in labour or consumption to the appropriate value of scaling factor (e.g. represent a labour shock of 10% as a labour scaling factor of 0.9).

ParameterShock objects must be passed to the main function shock_gtap as a Vector, which allows passing multiple shocks at once.

using EpiEconShocks

example_model = EpiEconShocks.Example.get_example_model();

# run the single function shock_gtap_example() applying a 5% labour shock,
# and a 20% shock to consumption of services
labour_shock = EpiEconShocks.ParameterShock("qe", ["skilled labor", "unskilled labor"], 0.05)
svces_shock = EpiEconShocks.ParameterShock("qpa", "svces", 0.2)

output = EpiEconShocks.Example.shock_gtap_example(
    example_model, [labour_shock, svces_shock])

# examine proportional change in GDP
comparisons = compare_gtaps(example_model, output)

comparisons.delta_gdp
7-element Named NamedArrays.NamedVector{Float64, Vector{Float64}, Tuple{OrderedCollections.OrderedDict{String, Int64}}}
A                  │ 
───────────────────┼──────────
oceania            │ -0.725887
asis               │  -0.74053
americas           │  -0.76541
eu                 │ -0.763694
other europe       │ -0.734144
mena               │ -0.632045
sub-saharan africa │ -0.624679

Further documentation

Further long-form documentation, a package index, and function documentation may be found at the package website.

Future development

This project is currently a work in progress. Future development may include the following:

  • More options for how GTAP endowments and commodities can be aggregated;

  • More fine-scaled control over the scaling applied to each sector within the GTAP regions, commodities, and endowments sets; for example, the ability to model a pandemic shock in a single region to examine global impacts, and similar for commodity demand.

  • Better documentation of how epidemic shocks can be modelled in more details, for example to include long-term impacts such as reductions in population size or increases in epidemic-related disability.

  • The Daedalus integrated epidemiological and macro-economic model, available as an R package daedalus. There is also a lagging Julia translation in the package Daedalus.jl.

  • EpiEconShocks.jl wraps the GlobalTradeAnalysisProjectModelV7.jl package which implements the GTAP version 7 trade model, which compatible with version 11 GTAP data (year 2017). Compatibility with other versions of the GTAP database is not assured.

Help

To report a bug, request a feature, or start a discussion, please open an issue.

References

WIP