An application that stores data about the busyness of NCSU's libraries and compiles it into an html summary using rmarkdown.
A Node/TypeScript/Express API for interacting with the database. Includes routes for creating records and getting records--either all records, n recent records, or records since a given datetime.
Handles the MySQL database. Contains one script other than the Dockerfile, schema.sql, that is run on container startup and creates tables if they do not exist.
When run, saves all recorded occupancy data to CSV files.
Contains functionality for knitting the rmarkdown file into an html document. Note that in development, using renv is preferred; however, while using docker, R packages are installed using apt to speed up image build time.
Handles an NGINX web server that serves the knitted rmarkdown html file.
Contains a python script that fetches current occupancy data and sends it to the TypeScript API.
Contains a python script that trains a timeseries forecasting model on the busyness data, then sends some predictions back to the database
Contains the React frontend for the predictions. Displays an interactive table and graph of future predictions.
Required environment variables:
MYSQL_ROOT_PASSWORD- the password to set for the root user in the db containerMYSQL_USER- the user to use in the data logger when connecting to the databaseMYSQL_PASSWORD- the password to use in the data logger when connecting to the database
Other environment variables and their defaults:
MYSQL_DATABASE=busyness- the name of the database to save records toMYSQL_HOST=db- the database hostnameAPI_HOST=api- the hostname for the api serverAPI_PORT=3000- the port of the api server
Start nginx server: docker compose up server -d
Run updater: docker compose run --rm updater
Run knitter: docker compose run --rm knitter
Backup data to CSV files: docker compose run --rm backup. NOTE: ensure that the export folder is present in the data/backup folder.
Run predictor model: docker compose run --rm predictor_model
Start predictor frontend: docker compose up predictor_frontend
Stop all services: docker compose down