relsa-severity-assessment - Assess animal study severity with RELSA
Computes reference-relative RELSA severity scores and exploratory ARIMA forecasts for laboratory animal study welfare measurements.
Tags
Updated: 2026-10-08Capabilities
Typical Inputs
Typical Outputs
What this skill does
- Compute RELSA severity scores
- Normalize measurements against baselines
- Build reusable reference sets
- Forecast scores with ARIMA
- Define KDE severity thresholds
- Evaluate forecasts with RMSE PICP MPIW
Inputs
- Animal measurement CSV
- Animal identifiers
- Time points
- Welfare measurements
- Variable directionality
- Baseline time
- Reference cohort
- Forecast target times
Outputs
- Per-animal severity scores
- Variable-level score weights
- ARIMA score forecasts
- Forecast prediction intervals
- Severity thresholds or zones
- Forecast evaluation metrics
- Reference model files
- Score output CSV files
Requirements
- Python 3.12 or newer
- numpy, pandas, scipy, statsmodels
- matplotlib for figures
- Filesystem read and write access
- Bash execution access
- Network access for installation or upstream review
