evo
Routes evo trajectory-evaluation workflows for APE/RPE, trajectory I/O, result analysis, configuration, and programmatic usage.
SKILL.md08-移动开发/evo/SKILL.md
GitHubevo
Use this skill for the evo Python package, which evaluates odometry and SLAM trajectories, loads and converts trajectory files, compares saved results, manages package settings, and supports programmatic plotting and notebooks.
Install
- Python: 3.10 or newer.
- Install from a checkout with
pip install -e .. - Install from PyPI with
pip install evo. - Optional extras only when you need them:
pip install evo[rerun]for Rerun workflows.pip install evo[gui]for the Qt plot backend.pip install evo[geo]for map tiles viacontextily.
Minimal checks
- Import check:
python -I -c "import evo; print(evo.__version__)" - Safe helper:
python scripts/check_env.py - CLI help:
evo --help
Route map
| User need | Go here |
|---|---|
| APE / RPE scoring, alignment, delta selection, result zips | sub-skills/metrics/SKILL.md |
| Trajectory file formats, converters, sync, align, export, bag / MCAP loading | sub-skills/trajectory-data/SKILL.md |
| Saved result comparison, merge, table export, label cleanup | sub-skills/result-analysis/SKILL.md |
evo_config, evo pkg, evo cat_log, evo_ipython |
sub-skills/configuration/SKILL.md |
Custom Python scripts, notebooks, plotting, Rerun, contextily |
sub-skills/python-api/SKILL.md |
When to read the bundled references
- Read
references/repo-provenance.mdwhen you need to know whether this skill still matches the current checkout. - Read
references/troubleshooting.mdfor cross-cutting install/import, optional dependency, backend, and settings problems. - Run
scripts/check_env.pywhen you want a quick import-and-CLI smoke check from an installed environment.
Working rules
- Use the sub-skill that matches the user’s primary workflow instead of mixing APE/RPE, trajectory I/O, result comparison, and configuration into one answer.
- Treat trajectory and metric helpers as in-place mutators unless the reference says otherwise; copy inputs when you need to preserve originals.
- Keep references and scripts self-contained inside this skill tree. Do not rely on the original checkout being present.
- If the package or APIs change, compare the checkout against
references/repo-provenance.mdand refresh the skill when needed.