analyze_region_heat
Use this skill when the user wants to compute regional activity heat, check-in volume, point density, unique users, or Top-K active geohash regions from trajectory records.
SKILL.mdskills/custom/spatiotemporal_trajectory/analyze_region_heat/SKILL.md
GitHubanalyze_region_heat
Use this Skill to aggregate point-level or evidence-level trajectory records into regional heat indicators.
Command
cd /mnt/skills/custom/analyze_region_heat
python3 scripts/analyze_region_heat.py \
--input /path/to/cleaned_points.jsonl \
--output-dir /path/to/output \
--geohash-precision 6 \
--time-bucket hour
Useful filters:
--time-start 2012-06-01T00:00:00 --time-end 2012-06-30T23:59:59
--hour-start 7 --hour-end 9 --geohash wx4
Outputs
region_heat.jsonlsummary.json
Algorithm
- Optional time/space filtering
- Geohash region aggregation
- Activity count
- Unique user count
- Optional category Top-N statistics
Next Skills
Use region_heat.jsonl with $detect_flow_anomaly for regional flow anomaly detection.