great-lakes-microplastics

Microplastics in the Great Lakes. Final website can be viewed here: (labs.waterdata.usgs.gov/visualizations/microplastics/index.html)[https://labs.waterdata.usgs.gov/visualizations/microplastics/index.html]

Visualization Preparation Pipeline

Data become figures by flowing along this path:

source -1-> data -2-> munge -3-> visualize

The functions implementing each arrow are in the scripts folder and/or in the VIZLAB R package. For example, the function used to implement -1-> is readData(), defined in scripts/lib/readData.R. The functions used to implement -2-> include mungeRelativeAbundance, defined in scripts/1_munge/mungeRelativeAbundance.R.

The overall flow is controlled by the remake::make() function, which uses three remake yaml files, each of which controls one of the above arrows. One arrow / remake yaml describes one processing stage for all products at once. The remake yaml files describe which products need to be made at that stage and which functions are required to make each product. The three yaml files are:

-1->: data.yaml (e.g., remake::make(remake_file='data.yaml'))

-2->: munge.yaml (e.g., remake::make(remake_file='munge.yaml'))

-3->: figures_R.yaml (non-R figures will be described elsewhere) (e.g., remake::make(remake_file='figures_R.yaml'))

-4->: layout.yaml (e.g., remake::make(remake_file='layout.yaml'))

Each remake script also points backward to the preceding remake script, which means that a single call to remake::make(remake_file='figures_R.yaml') is enough to ensure that the entire pipeline is up to date.

To start a simple local server, navigate to the target folder in a shell, and run the command:

python -m SimpleHTTPServer

Ctrl-C quits the server.

Running with Docker

There is a Dockerfile and Makefile to run the above steps all in one. It sets up the environment with all packages and dependencies needed. To run this type the following:

docker build -t great-lakes-microplastics .
docker run -v $(pwd)/target:/opt/viz/target -v $(pwd)/cache:/opt/viz/cache -u $UID great-lakes-microplastics

This will dump the target outputs to your target directory and you will still need to run the python SimpleHTTPServer against it.