Statistical Tools for Modelling Climate-Health Impacts.

Climatehealth Package

Table of Contents

• Package Overview
• SOSCHI Topics and Indicators
• Package Installation
• Running analysis
• Contributing to the package
• Authors
• Contributors
• Project Lead
• Methodology review
• Acknowledgements
• Copyright
• Citation
• Funding

Package Overview

Description

This package provides functions for calculating climate–health indicators and supporting the production of climate–health statistics. The methods implemented in the package follow the statistical framework for climate–health indicators developed under the Standards for Official Statistics on Climate–Health Interactions (SOSCHI) project. You can find out more and stay up to date with the SOSCHI project here.

The methods for measuring headline outcome indicators for the SOSCHI topics were endorsed by the 57th United Nations Statistical Commission for inclusion in the Global Set of Climate Change Statistics and Indicators (see decision 57/109 (b) in the Final Report).

The package is also now available on CRAN.

SOSCHI Topics and Indicators

This package contains scientific functions dedicated to the following topics in the SOSCHI framework. Please note the final methods documents are still in development, so the versions linked below are earlier alpha versions.

[!NOTE] Please note that there may be some inconsistencies between the indicators and methods included in the Github code/R package, and those described in the corresponding SOSCHI topic documents, and/or related analysis publications. This is because the final methods documents are still in development, so those linked below in the section 'SOSCHI Topics and Indicators' are an earlier alpha version.

For information on how to use this R package, check here

Package Installation

Requirements

To install and use the package, you need an R environment. This package has been developed primarily with R 4.4.1, and that version is recommended where possible. R-4.4.1 can be downloaded here

Installation options

You can use the package in several ways, depending on whether you want a stable release or want to work with the source code.

1. Install the stable CRAN release

If you want the latest stable public release, install the package directly from CRAN:

install.packages("climatehealth")

You can also browse the CRAN package page here.

2. Load from a local GitHub clone for development

Use this option if you want to inspect the source code, make changes locally, test updates, or contribute to the package.

First clone the repository from GitHub to your machine. Then in R:

install.packages("devtools")
devtools::load_all(path = "{path/to/climatehealth}")

This loads the package directly from the local source directory into your current R session. This is useful for development, but won't make the climatehealth package available as a standard installed package in your library. Once the source code changes, run devtools::load_all() again to reload the updated functions. If you are an end user, follow the standard installation procedure outlined below for your operating system.

3. Install directly from GitHub

If you want to install the latest development version of the package directly from GitHub without cloning the repository, you can use remotes::install_github().

First install the remotes package in R if needed:

install.packages("remotes")

Then install the package with:

remotes::install_github("onssoschi/climatehealth")

This installs the package from the GitHub repository into your R library in the usual way. This option is useful if you want the latest development version of the package but do not need to edit the source code locally.

4. Install the latest Windows binary release (pre-compiled)

If you would like to install the latest official release of the package, download the .zip file from the latest GitHub release here. Then install it in R with:

install.packages(path = "{path/to/climatehealth_<version>.zip}", repos = NULL, type = "win.binary")

where <version> is the version number of the release you downloaded.

5. Install the latest source release for MacOS or other platforms (needs compilation)

If you are using MacOS and would like to install from the source, download the .tar.gz file from the latest GitHub release here. Then install it in R with:

install.packages(path = "{path/to/climatehealth_<version>.tar.gz}", repos = NULL, type = "source")

where <version> is the version number of the release you downloaded.

Running analysis

Once the package is installed or loaded, you can run any of the main analysis entry points depending on your use case.

Indicator workflows

Most indicator modules are run through a top-level *_do_analysis() function. For example:

result <- climatehealth::air_pollution_do_analysis(
  data_path = "{path/to/your_data.csv}"
)

Other indicators follow the same pattern, for example:

temp_mortality_do_analysis()
wildfire_do_analysis()
malaria_do_analysis()
diarrhea_do_analysis()
suicides_heat_do_analysis()

Each analysis function expects specific input data and parameters for that indicator, so refer to the package documentation for the relevant module before running it.

Example usage scripts for the main indicator workflows are provided in inst/examples/, including:

• air_pollution_usage.R
• temp_mortality_usage.R
• wildfires_usage.R
• malaria_usage.R
• diarrhea_usage.R
• suicides_heat_usage.R
• descriptive_stats_usage.R

Descriptive statistics

You can also run the descriptive statistics workflow directly:

result <- climatehealth::run_descriptive_stats(
  data = my_data,
  output_path = "{path/to/output}"
)

This is useful for checking data quality, exploring variables, and generating summary outputs before running an indicator-specific workflow.

Contributing to the package

We welcome contributions from public users, collaborators, and developers who want to improve the package. If you want to contribute code, documentation, tests, or bug fixes, please start by reading the CONTRIBUTING guide CONTRIBUTING. This includes guidance on setting up a local development workflow, making changes safely, and submitting updates through GitHub.

Authors

The package authors reflect direct intellectual and technical contributions to the package architecture, module methods, and reproducible analytical workflows.

• Lead author: Charlie Browning
• Lead maintainer and platform developer: Kenechi Omeke
• Authors: Etse Yawo Dzakpa, Gladin Jose, Matt Pearce, Ellie Watkins, Claire Hunt, Beatrice Byukusenge, Cassien Habyarimana, Venuste Nyagahakwa, Felix Scarbrough, Treesa Shaji, Bonnie Lewis, and Vijendra Ingole

Contributors

The following contributors supported the package through bug fixes, minor features, testing, documentation, and review work:

• Sean Lovell
• Antony Brown
• Euan Soutter
• Gillian Flower
• David Furley
• Joe Panes
• Charlotte Romaniuk
• Milly Powell

Project Lead

• Myer Glickman, Bonnie Lewis and Vijendra Ingole

Methodology review

• Sewe Maquines

Acknowledgements

We acknowledge the Topic Expert Group members and Expert Advisory Group members for higher-level leadership, methodological input, and subject-area support to the SOSCHI climate-health work. We also acknowledge the use of external data and code examples that informed testing and early prototyping.

Copyright

This is an open access R package published under a UK Open Government Licence 3.0 which is compatible with the Creative Commons CC-BY-SA 4.0 licence. You are free to copy, publish, distribute and adapt the information, provided you acknowledge the source and link to this R package (see citation below). Enquiries concerning the application of this R package can be made to the project team at [email protected].

Citation

If you use climatehealth in publications, cite the package as:

Browning, C., Omeke, K., Dzakpa, E. Y., Jose, G., Pearce, M., Watkins, E., Hunt, C., Byukusenge, B., Habyarimana, C., Nyagahakwa, V., Scarbrough, F., Shaji, T., Lewis, B., and Ingole, V. (2026). climatehealth: Statistical Tools for Modelling Climate-Health Impacts. R package.

An R package citation file is provided at inst/CITATION.

Funding

This work was supported by Wellcome.