Installation

Before you start

Install Julia with juliaup, the official installer. EpiBranch needs Julia 1.10 or later, and the household and network packages need Julia 1.11 or later. Start Julia by typing julia in a terminal, or use an editor such as VS Code with the Julia extension.

Julia keeps the packages for each analysis in a project folder, much as renv does for R. The project records which package versions you used in two files, Project.toml and Manifest.toml. Keep them with your analysis so it can be rerun later with the same versions.

The first time you load EpiBranch, and the first time you call a function in a new session, Julia compiles code, which can take from a few seconds to a minute. Later calls in the same session run at full speed.

You do not need R to use EpiBranch. Julia for R users explains the syntax used in these pages.

The core package only

EpiBranch is in the General registry, Julia's equivalent of CRAN, but the registered version is behind these pages. Install the development version from GitHub for now. You load it with using EpiBranch, as you would library() a package in R:

using Pkg
Pkg.activate("my-analysis")  # a project for this analysis, as above
Pkg.add(
    PackageSpec(url = "https://github.com/epiforecasts/EpiBranch.jl", rev = "main")
)
# The examples in these pages also use these
Pkg.add(["Distributions", "StableRNGs", "DataFrames"])

using EpiBranch

A few later pages need more, such as Graphs for the network models and Turing for the pages on fitting. The using lines at the top of a page name what that page needs.

The registered version is older than these pages

The registered version (0.1.0) does not run the examples in these pages, which describe version 0.2. Setting how long isolation or quarantine lasts (duration), for example, fails with an error. The block above installs the development version from GitHub for that reason; Pkg.add("EpiBranch") would fetch 0.1.0.

With the household and network packages

The household and network packages (EpiHouseholds and EpiNetwork) need a newer version of EpiBranch (0.2) than the one registered (0.1.0). Until 0.2 is released, install all three from the development version on GitHub (the main branch):

using Pkg
Pkg.activate("my-analysis")
repo = "https://github.com/epiforecasts/EpiBranch.jl"
# The development version, which these pages describe
rev = "main"
Pkg.add([
    PackageSpec(url = repo, rev = rev),
    PackageSpec(url = repo, rev = rev, subdir = "lib/EpiHouseholds"),
    PackageSpec(url = repo, rev = rev, subdir = "lib/EpiNetwork"),
])
# The examples in these pages also use these
Pkg.add(["Distributions", "StableRNGs", "DataFrames"])

PackageSpec says where to install a package from: here the GitHub repository, a branch or commit (rev), and for the two companion packages the folder within the repository that holds them (subdir). Running Pkg.update() later brings all three up to the latest development version. To fix an analysis to one version instead, set rev to the full identifier (hash) of a commit from the commit list on GitHub, the same one for all three packages.

Once EpiBranch 0.2 is released this becomes an ordinary Pkg.add.