tiemvanderdeure/AfricanMalariaVectors: v1.0

Code for "Climate change favours African malaria vectors" This repository contains code for "Climate change favours African malaria vectors" by van der Deure, Nogués-Bravo, Njotto & Stensgaard, which is currently in preparation. In this publication, we investigate the current and future climate and land use conditions for six dominant and widely spread Africa malaria vectors. All code used in the analysis is contained in this repository, and all data is either automatically downloaded when the scripts are run, or can be downloaded using the instructions below (the code will prompt when the file is not found in the expected location). Understanding this repository This repository is formatted as a Julia package. The Project.toml specifies which packages it relies on. The src folder contains its internal code and the data folder contain data files. The only code you have to interact with directly to reproduce all analysis in the paper is located in the scripts folder Reproducing the analysis First clone this repository and navigate to the folder. Installing julia and instantiating the environment All analysis is performed in the Julia 1.11. The recommended way to install Julia is through juliaup. After installing julia, navigate to this repository, then type ] in the Julia REPL to enter package mode, and type activate . followed by instantiate build a reproducible environment that contains all packages necessary to run the code. This project uses R to fit GAMs and to download Malaria Atlas Project data, interfacing between the languages using RCall.jl. A working R installation is required and will be automatically detected and the necessary R packages installed. R version 4.4.2 was used to run the code. Downloading data Most of the data will be automatically downloaded. This includes large amounts (10s of GB) of raster data, which will be downloaded to a path specified in ENV["RASTERDATASOURCES_PATH]. You can set this to any path (e.g. a hard-drive) by running ENV["RASTERDATASOURCES_PATH"] = "my/data/path" in the Julia REPL. The data will then be stored in several sub-folders. Vector and malaria observation data is read from the data folder. Two sources cannot be directly downloaded and need to be downloaded manually: The geo-coded inventory of anophelines by Snow (2017). It is available at doi.org and contains a .csv file that should be copied to data, so it can be loaded from data\Africa Vectors database_1898-2016.csv. The Plasmodium falciparum prevalence datast by Snow (2017). It is available at doi.org and contains a .csv file that should be copied to data, so it can be loaded from data\00 Africa 1900-2015 SSA PR database (260617).csv. Running the code The scripts folder contains three Julia scripts: scripts/main.jl runs the analysis, while scripts/figures.jl and scripts/tables.jl use the Julia objects to create figures and tables in the publications. These are saved in the images and tables folders. You could run these from the command line by navigating the this folder, opening a Julia REPL by running the julia command, and then running each of the three the scripts using an include statements (include("scripts/main.jl") etc). Alternatively, you can use an environment like Visual Studio Code to run them line-by-line. This code makes use of multi-threading to speed up computation, see the relevant section of the Julia manual for how to run Julia with multiple threads. Contact For any questions about the code in this repository, contact Tiem van der Deure at email tvd@sund.ku.dk. Making an issue in the GitHub repository works as well.