Prayag Gordy, Ankit Patel, Hallie Trial
To run our code, either source main.R or knit paper.Rmd.
main.R first loads the requisite libraries, then sources the R scripts in the R/ folder (see line 3, sapply(list.files(path = "R", full.names = T), source, .GlobalEnv)). Each of these scripts contains only functions, which are then loaded into the environment. These functions are later used in main.R to create objects, which are used in further functions. Finally, we create our SQL database.
In functions that load data, you'll see an update parameter whose default value is false. If set to true, the function will pull the data again from the original source; if false, the function will check for an already-processed copy of the data or at least an already-downloaded raw copy, and it will use that instead of downloading again. The first time you run the script, all data will have to download; this will take a little while!
paper.Rmd first sources main.R to load everything into the environment. Then it knits the children documents—one for each section of the paper—found in the Rmd/ folder. These documents are in the same environment as paper.Rmd and therefore have access to all the variables and functions loaded in main.R.
Almost every single function in this repository makes use of our config.yaml file. In our config file, we store paths, filenames, URLs, and other hard-coded values in a YAML format. This strengthens our code by holding all hard-coded values in one location instead of spread throughout a large codebase.