Prepare Source Data

Firecube expects you to provide source datasets. In normal use, these files may come from a data provider, an existing local source, or object storage. You are responsible for staging source files that match what your installed plugin accepts, then passing their directory or storage prefix to --input-data. The installed plugin decides which formats, filenames, and dataset structure it accepts; nothing about that is generic.

The quickstart_plugin plugin from the previous page expects time-indexed NetCDF files. It ships a script that generates four matching example files, so this quickstart does not require you to source real data.

Create The Example Files

Run this command from the firecube-quickstart-plugin/ directory, with the quickstart virtual environment active:

python scripts/generate_sample_data.py

Verify The Source Data

List the generated files:

ls sample_data

Expected output:

sample01.nc  sample02.nc  sample03.nc  sample04.nc

Each file contains one timestamp of temperature_c and humidity_pct on the same 2 by 3 latitude-longitude grid. The ingestion command passes their parent directory to the plugin.

Next Steps

• Run Ingestion: convert the four NetCDF files into one local Zarr product and inspect it.