Python tools for managing data in the Lightcurve Database. This database stores fluxes as a function of time for tracked objects. This package includes simulation tools to fill the database with test data.
Installation - first download the git repository, then
uv pip install -e .
You can spin up an ephemeral instance of the database with the
lightcurvedb-ephemeral
tool. This uses testcontainers. It will print connection information.
The ephemeral command also creates deterministic fixtures for unassigned-source
cross-matching. The default set includes fixed catalogue-like and unmatched
sources, a moving source, and a nearby pair for merge review. The fixture count
must be at least six to retain every scenario; use --unassigned-seed to
reproduce the simulated source values. Fixture IDs use UUID7. For a fast,
unassigned-only Parquet fixture
database:
lightcurvedb-ephemeral --backend parquet --number 0 --unassigned-number 6 \
--unassigned-measurements 10 --unassigned-seed 20260722