byMark Tiele Westra via web
As preparation to the online Liberia water portal, I analysed the most recent water point survey data, that was collected at the beginning of this year. About 20.000 datapoints were collected all over Liberia, capturing location, water point type, functionality, usage, etc. As the final report is still under review — it will be shared on http://wash-liberia.org when ready — I thought I'd share a few words on which tools were used to do the analysis.
We wanted our analysis to be reproducible (https://en.wikipedia.org/wiki/Reproducibility#Reproducible_research), so anyone with the required skills can check our work, or learn from it. Therefore, we chose to use the popular, open source and freely available statistical computation and visualisation language R (https://www.r-project.org/) to analyse the data. However, this basically gives you a bunch of numbers and graphs.
To turn these results into a nice report, we used R Markdown (http://rmarkdown.rstudio.com/), which is a simple, text-file based markup language that can be used to create documents, including all the headings, subheadings, and figure captions that you might want. Importantly, the R markdown includes the results of the analysis.
Finally, using the R library knitr (https://yihui.name/knitr/), the whole thing is turned into a nice PDF or Microsoft Word document. The end result can even be styled to your heart's content. And if your underlying data changes? Just click 'Run' again, and your whole document is recreated from scratch. Nice!
When the review process is done, we plan to publish the raw data plus the R and R Markdown code, so anyone can take a closer look.