README.Rmd

---
output: github_document
---

<!-- README.md is generated from README.Rmd. Please edit that file -->

```{r, include = FALSE}
knitr::opts_chunk$set(
  collapse = TRUE,
  comment = "#>",
  fig.path = "man/figures/README-",
  out.width = "100%"
)
```

# An open-source framework for Rift Valley Fever forecasting

<!-- badges: start -->
[![Project Status: WIP – Initial development is in progress, but there has not yet been a stable, usable release suitable for the public.](https://www.repostatus.org/badges/latest/wip.svg)](https://www.repostatus.org/#wip)
[![Lifecycle: experimental](https://img.shields.io/badge/lifecycle-experimental-orange.svg)](https://lifecycle.r-lib.org/articles/stages.html#experimental)
[![License (for code): MIT](https://img.shields.io/badge/License (for code)-MIT-green.svg)](https://opensource.org/licenses/MIT)
[![License: CC0-1.0](https://img.shields.io/badge/License (for data)-CC0_1.0-lightgrey.svg)](http://creativecommons.org/publicdomain/zero/1.0/)
[![License: CC-BY-4.0](https://img.shields.io/badge/License (for text)-CC_BY_4.0-blue.svg)](http://creativecommons.org/publicdomain/zero/1.0/)
<!-- badges: end -->

### OpenRVFcast
EcoHealth Alliance's ongoing OpenRVFcast project is developing a generalizable, open-source modeling framework for predicting Rift Valley Fever (RVF) outbreaks in Africa, funded by the Wellcome Trust's climate-sensitive infectious disease [modeling initiative](https://wellcome.org/news/digital-tools-climate-sensitive-infectious-disease). We aim to integrate open data sets of climatic and vegetation data with internationally-reported outbreak data to build an modeling pipeline that can be adapted to varying local conditions in RVF-prone regions across the continent.

### Repository Structure and Reproducibility
-   `data/` contains downloaded and transformed data sources. These data are .gitignored and non-EHA users will need to download the data. 
-   `R/` contains functions used in this analysis. 
-   `reports/` contains literate code for  R Markdown reports generated in the analysis
-   `outputs/` contains compiled reports and figures.
-   This project uses the [{renv}](https://rstudio.github.io/renv/) framework to 
    record R package dependencies and versions. Packages and versions used are recorded in `renv.lock` and code used to manage dependencies is in `renv/` and other files in the root project directory.  On starting an R session in the working directory, run `renv::restore()` to install R package dependencies.
-   This project uses the [{targets}](https://wlandau.github.io/targets-manual/) 
    framework to organize build steps for analysis pipeline. The schematic figure below summarizes the steps. (The figure is generated using `mermaid.js` syntax and should display as a graph on GitHub.  It can also be viewed by pasting the code into <https://mermaid.live>.)
    
    
```{r, echo=FALSE, message = FALSE, results='asis'}
mer <- targets::tar_mermaid(targets_only = TRUE, outdated = FALSE, 
                            legend = FALSE, color = FALSE, 
                            exclude = c("readme", ends_with("_targets")))
cat(
  "```mermaid",
  mer[1], 
  #'Objects([""Objects""]) --- Functions>""Functions""]',
  'subgraph Project Workflow',
  mer[3:length(mer)],
  'linkStyle 0 stroke-width:0px;',
  "```",
  sep = "\n"
)
```

Many of the computational steps can be time consuming and either depend on or produce large files. In order to speed up the pipeline, intermediate files can be stored on the cloud for rapid retrieval and portability between pipeline instances. We currently use an AWS [S3 bucket](https://aws.amazon.com/s3/) for this purpose. The pipeline will still run without access to cloud storage but the user can benefit from adapt the `_targets.R` file to use their own object storage repository. AWS access keys and bucket ID are stored in the `.env` file. 

For handling gridded binary weather data, this pipeline uses the [ecCodes](https://confluence.ecmwf.int/display/ECC) package which can be installed on OSX using `homebrew`:

```
brew install eccodes
```

and on Ubuntu using apt-get

```
sudo apt update
sudo apt install eccodes
```

EHA users: see the `stripts/` repository to be able to directly download data from AWS outside of the targets workflow. 

Follow the links for more information about:

- [`targets`](https://ecohealthalliance.github.io/eha-ma-handbook/3-projects.html#targets)
- [`renv`](https://ecohealthalliance.github.io/eha-ma-handbook/3-projects.html#package-management-with-renv)  
- [git-crypt](https://ecohealthalliance.github.io/eha-ma-handbook/16-encryption.html)
- [Reproducible workflows](https://github.com/ecohealthalliance/building-blocks-of-reproducibility)