Oxford International Health and Tropical Medicine Hackathon 2024

This repository contains instructions, data and code for the University of Oxford MSc in International Health and Tropical Medicine Hackathon 2024. The Hackathon 2024 event is part of the MSc course’s lecture series on Open Science and Reproducible Research in R.

Motivation

Hackathon 2024 caps the students’ introduction to Open Science and Reproducible Research in R through an actual global health project within which they are to serve as researchers/data scientists. This exercise aims to provide the students a platform from which to apply skills in working with data using R that they have been learning and practising for about the past 6 weeks while at the same time exposing them to a collaborative team environment.

Format

Hackathon 2024 is structured as a problem-based learning exercise, a format that the course students are already familiar with given similar approaches done for other lectures. Briefly, this PBL exercise presents the problem first rather than teaching relevant material and subsequently having students apply the knowledge to solve the problem. Whilst the previous lectures in the Open Science and Reproducible Research in R series have provided foundational skills in R, the PBL approach for this hackathon will challenge the students to further explore and learn the extensive functionalities R has to offer in order to appropriately solve the problem/s they have been given to solve. This PBL is group-orientated and simulates a collaborative research/data science working environment facilitated through the use of git and GitHub.

Through this approach, the students are expected to:

1. Examine and define the problem.

2. Explore what they already know about underlying issues related to it.

3. Determine what they need to learn and where they can acquire the information and tools necessary to solve the problem.

4. Evaluate possible ways to solve the problem.

5. Solve the problem.

6. Report on their findings.

The case

The case study along with the hackathon rules are presented here - https://oxford-ihtm.io/ihtm-hackathon-2024/case_study.html.

Repository/project structure

This R project has the following structure:

ihtm-hackathon-2024
    |-- data/
    |-- docs
      |-- case_study.html
    |-- outputs/
    |-- packages.R
    |-- R/
    |-- reports
      |-- case_study.Rmd
      |-- sudan_health_nutrition.Rmd
    |-- sudan_health_nutrition_1.R
    |-- sudan_health_nutrition_2.R
    |-- sudan_health_nutrition_3.R
    |-- sudan_health_nutrition_4.R
    |-- sudan_health_nutrition_5.R
    |-- sudan_health_nutrition_6.R
    |-- sudan_health_nutrition.R

• data/ contains the datasets required for this project. These datasets may either be provided by the module leader and/or retrieved programmatically by the hackathon participants and saved into this directly;

• docs/ contains compiled reports produced from Rmarkdown documents;

  Currently, this folder contains the file case_study.html which is the HTML output of case_study.Rmd (see below) which details the case study and the rules for this hackathon;

• outputs/ contains figures, intermediate or final results tables, and other outputs produced by this scientific workflow;

• packages.R is the overall R script that declares the R package dependencies of the entire project;

• R/ contains functions created or that will be created for use in this project;

  Currently, this folder contains an R file called get_data.R. This file specifies a function called get_data() which retrieves and processes the data for this hackathon. It also saves this dataset as a CSV file called nutrition_survey_dataset.csv in the data folder;

• reports/ contains Rmarkdown file/s for producing report/s produced by this scientific workflow;

  Currently, this directory contains the following files

  • case_study.Rmd - Rmd file that produces the case_study.html file found in the docs folder
  • sudan_health_nutrition.Rmd - Rmd file for producing the overall report

• sudan_health_nutrition_1.R is the R workflow script for section 1 of the analysis;

• sudan_health_nutrition_2.R is the R workflow script for section 3 of the analysis;

• sudan_health_nutrition_3.R is the R workflow script for section 3 of the analysis;

• sudan_health_nutrition_4.R is the R workflow script for section 4 of the analysis;

• sudan_health_nutrition_5.R is the R workflow script for section 5 of the analysis;

• sudan_health_nutrition_6.R is the R workflow script for section 6 of the analysis; and,

• sudan_health_nutrition.R is the overall R workflow script for the complete analysis and reporting.

Reproducibility

This project is built on R version 4.3.2.

To work on this project, please follow these steps:

1. Clone this project onto your local machine. Instructions on how this is done can be found here.

2. In your local clone of the project, please make sure to create a new branch from the main branch. You should name this branch in such a way that uniquely identifies it as your personal branch (i.e., give it your name). Please avoid blank spaces in branch names. If you need to put a space, use a - or a _.

3. Install all declared R package dependencies found in packages.R.

   • Check if the packages listed in packages.R are already installed using the following code:

   installed.packages() |>
     (\(x) x[ , 1])() |>
     (\(x) x[c("name_of_package1", "name_of_package2", "name_of_package3")])()

   Please run this code direct to your R console rather than encoding in the R scripts in this project.

   Please make sure to replace the placeholder text name_of_package1 etc with the actual names of the packages listed in the packages.R file.

   This code will show which of the packages listed in packages.R are already installed in your computer. Packages listed in packages.R that are not shown in the output of the code above are the packages that are not yet installed in your computer.

   • Install the packages that are not yet installed using the following code:

   install.packages(c("name_of_package1", "name_of_package2", "name_of_package3"))

   Please run this code direct to your R console rather than encoding in the R scripts in this project.

   Please make sure to replace the placeholder text name_of_package1 etc with the actual names of the packages listed in the packages.R file that are not yet installed in your computer.

Once all R package dependencies have been installed, you should now be able to work on this project on your own branch and make changes/contributions as directed by project lead.

Running the workflow

Running the entire workflow

To run the entire workflow, issue the following command onto R console:

source("sudan_health_nutrition.R")

Running specific sections of the workflow

The project workflow is currently divided into 6 discrete processes implemented in 6 different R scripts labelled:

• sudan_health_nutrition_1.R is the R workflow script for performing bottleneck analysis of antenatal care for pregnant women;

• sudan_health_nutrition_2.R is the R workflow script for performing bottleneck analysis of expanded programme on immunisation (EPI) or children;

• sudan_health_nutrition_3.R is the R workflow script for performing barriers to education access analysis;

• sudan_health_nutrition_4.R is the R workflow script for performing spatial distribution analysis of child and maternal undernutrition;

• sudan_health_nutrition_5.R is the R workflow script for performing CMAM programme responsiveness analysis; and,

• sudan_health_nutrition_6.R is the R workflow script for determinants of child and maternal undernutrition analysis.

To run any of these, issue the following commands in the R console:

# Setup the workflow environmen ----

## Load packages in packages.R and project-specific functions in R folder ---- 
suppressPackageStartupMessages(source("packages.R"))
for (f in list.files(here::here("R"), full.names = TRUE)) source (f)


## Read data ----
maternal <- read.csv("data/maternal_health.csv")
child <- read.csv("data/child_health.csv")
cmam <- read.csv("data/cmam_routine_data.csv")

### Retrieve and read Sudan map data ----
sudan_map_spec <- download_sudan_maps(download_url = "https://data.humdata.org/dataset/a66a4b6c-92de-4507-9546-aa1900474180/resource/e5ef3cc7-f105-4565-8d73-e08bb756f1c1/download/sdn_adm_cbs_nic_ssa_20200831.gdb.zip")
sudan_map_url <- "https://github.com/spatialworks/sudan/raw/master/data-raw/maps/sudan.gpkg"

sudan0 <- st_read(dsn = sudan_map_spec$dsn, layer = sudan_map_spec$layers[1])
sudan1 <- st_read(dsn = sudan_map_url, layer = "state")
sudan2 <- st_read(dsn = sudan_map_url, layer = "locality")

## Run the specific workflow ---
source("sudan_health_nutrition_1.R")

Reproducing the final report for this project

To reproduce the HTML final report for this project, run the following command on the R console:

rmarkdown::render(
  "reports/sudan_health_nutrition.Rmd", 
  output_dir = "docs", 
  knit_root_dir = here::here()
)

This will render the Rmarkdown found in the reports directory and produce an HTML report called “sudan_health_nutrition.html” in the docs directory.

Authors

• Anita Amponsa
• Shylett Anthony
• Naemi Araya
• Moshood Audu
• Rasika Bombatkar
• Neira Budiono
• Phillip Chigiya
• Prince Kelechi Chima
• John Bok Chol
• Clifford Cofie
• Eslam Elbasheer
• Prateek Garg
• Mary Gouws
• Samvel Grigoryan
• Marietta Imadojiemu
• Jillian Francise Lee
• Anita Makori
• Joseph Mwaka
• Kapil Narain
• Josephine Ndawula
• Thokozani Nyasulu
• Richmonda Pearce
• Mariano Ratto
• Gloria Rukomeza
• Mercedes Rumi
• Amina Suveha
• Shih-Ting Tseng
• Claudia Vidal Cuellar
• Yih Seong Wong

License

Unless otherwise specified, data used in this repository are licensed under a CC0 1.0 Universal license.

All code in this repository are licensed under a GNU General Public License 3 (GPL-3) license.