Skip to contents

Latest changes to malariaverse

Latest-changes.Rmd

📢 Latest update

This page describes a significant update to the malariaverse site files and several of the supporting packages. These changes affect — and in some cases break — existing workflows, so please read on if you have analyses built against the previous site files and packages.

⚙️ What’s changed

The malariaverse packages and site files have been substantially updated (full detail in site PR #56 and postie PR #8). The headline changes include:

  • 🌍 Population and demography methods updated to the latest WorldPop outputs.
  • 🌧️ Monthly rainfall estimates updated to CHIRPS v3.0.
  • 🗺️ Updated to the latest MAP rasters.
  • 📊 WHO figures updated to the 2025 World Malaria Report.
  • 💉 Updated UNICEF malaria vaccine dose-delivery data.
  • ✨ New functionality: the option to specify a parameter draw, a Plasmodium vivax model, and within-year (sub-year) intervention specification.
  • 🛏️ Refined intervention implementation — net modelling now aligns more closely with the MAP net model, IRS trial-to-implementation downscaling, and corrected SMC/PMC round timing.
  • ⚠️ itn_model_dist is now a required input (with helper functions and a detailed error message to guide you if it’s missing).
  • 🏥 The previous fixed shortcut (malaria deaths taken as a constant multiple of hospitalised severe incidence) is replaced by an explicit calculation that splits severe cases into hospital and community streams, each with its own case-fatality ratio.
  • 🏥 In postie, a new ft_sev input lets you set and vary over time the proportion of severe cases that reach hospital. It defaults to 0.8, which reproduces the previous behaviour.
  • 🧾 New disaggregated output columns (hospital vs community severe streams) and case-fatality-ratio arguments are available.

Importantly, alongside these data and methods updates, the structure of the site files themselves has changed, and the supporting malariaverse packages have been updated in step. Taken together, this means code written against the previous site files and packages will likely not run unchanged against the updated versions.

🔁 Reproducing existing work

Many of you will need to re-run or reproduce earlier analyses. Two things make that possible:

  1. Older site files remain available through fetch_site(), so you can continue to retrieve the previous data.
  2. We have frozen the previous versions of the malariaverse packages as a coordinated snapshot under the tag malariaverse_01_2025. Pairing the older site files with this tagged set of packages reproduces the environment your existing workflows were built for.

To install the frozen package set, e.g. with pak:

pak::pak(c(
  "mrc-ide/site@malariaverse_01_2025",
  "mrc-ide/postie@malariaverse_01_2025",
  "mrc-ide/[other-pkg]@malariaverse_01_2025"
))

Remember, you can see all available site-files and versions by calling site::available_sites()

🗓️ Package versions

These changes are reflected in the latest versions of the malariaverse packages and site files. If you have analyses that rely on the previous behaviour, we’d recommend noting which package versions you were using (or installing the tagged snapshot above) before updating.

❓ Questions

If you have any questions or run into problems, please open an issue on the site package issue page or contact us at .

🙏 Huge thanks to the many individuals who have contributed to these changes.