You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Sibo Wang-Chen edited this page May 26, 2026
·
5 revisions
The Spotlight Python tools are in the separate spotlight-tools repository. They implement:
Arena registration model fitting
Full postprocessing pipeline (behavior alignment, muscle warping, visualization)
We use uv for package management (previously Poetry). uv is significantly faster and still uses pyproject.toml.
Installation
# Install uv if not already installed (see https://docs.astral.sh/uv/)# Clone and install spotlight-toolscd spotlight-tools/
uv sync
Running scripts
After installation, entry points are available directly in your shell (or via uv run <command>):
# Fit the arena registration model
fit-arena-registration -a ~/Spotlight/arenas/arena146
# Post-process a recording
postprocess-recording --recording-dir ~/data/spotlight/20250613-fly1b-002/
# Visualize stage trajectory
visualize-stage-trajectory --behavior-frame-metadata-path .../processed/behavior_frames_metadata.csv --output-path .../stage_trajectory.png
Entry points are defined in the [project.scripts] section of spotlight-tools/pyproject.toml.
For maintainers
Update dependencies by editing pyproject.toml version bounds and running: