Shareable Timeseries Visualisation

The python-client can be used to upload data to a MySQL instance. The website can be used to inspect all of this data in real-time.

⚙️ How to use the python-client to upload data

1. Get your own database and credentials from Moritz Makowski

2. Put the credentials in a file config.json (format of config.example.json)

3. Fill the database using the STVClient (see /client/main.py)

from stv_client import STVClient

client = STVClient(
    database_name="your_database_name",
    table_name="any_table_name",
    data_columns=["ch4", "co2"],
    units={"ch4": "ppb", "co2": "ppm"},
    descriptions={"ch4": "The cows are at fault, for sure"},
    minimums={"ch4": 0, "co2": 0},
    decimal_places={"ch4": 3, "co2": 1},
    print_stuff=True,
)

client.insert_data("sensor-1", {"ch4": 1.815, "co2": 418.0})
client.insert_data("sensor-2", {"ch4": 1.845, "co2": 415.0})
client.insert_data("sensor-3", {"ch4": 1.900, "co2": 420.0})

print("latest 10 records: ", client.get_latest_n_records(10))

# Don't forget this line, it will close the SQL connection
del client

4. View and share the plots on the website (our instance: https://tueiesm-stv-api.netlify.app/)

5. Properly embed the client into your codebase using Git Submodules. An example can be found here: https://github.com/tum-esm/airquality_station/tree/WebUI-extended. /no-gui.py makes use of the stv-client.

🏛 Architecture

I. Upload data from the python-client

Tables will only be created/altered when the schema is new or changes.

II. Discover available database tables

Triggered on every website load.

III. Fetch data and meta-data for a given database and table

/meta-data fetched on every table-select, /data fetched on every table-select, and click of the refresh button.

🔒 Security

All of the uploaded data is publicly available.

The python-client uses a MySQL user, that has all privileges on all tables in the respective database and no privileges on any other databases.

The backend uses a MySQL user, that is allowed to read anything from all databases associated with this project.

When fetching data or meta-data, we don't have to look out for SQL injections since the backend only has read permissions on the public tables anyways - no other permissions that could be exploited.

💾 Setting up your own MySQL instance

/backend/README.md includes: 1. Example responses from the backend, 2. Commands how to set up the MySQL instance.

tum-esm/shareable-timeseries-visualization