Quickstart

This guide provides a quick introduction to the LoS Estimator using a synthetic dataset.

Installation

First, clone the repository and set up a virtual environment:

git clone git@git.rwth-aachen.de:jrc-combine/los-estimator.git
cd los-estimator
python -m venv .venv

# On Windows
.\.venv\Scripts\activate

# On Linux/macOS
source .venv/bin/activate

pip install -r requirements.txt

Synthetic Example

The package includes a demonstration script that generates synthetic ICU data and performs LoS estimation:

python examples/synthetic_example.py

This script demonstrates the complete workflow:

  1. Generates synthetic ICU admission time series

  2. Simulates occupancy data using a known lognormal LoS distribution

  3. Estimates the LoS distribution from the simulated data

  4. Compares estimated distributions against the ground truth

  5. Generates visualizations and animations

The configuration is read from examples/synthetic_example.toml. You can modify parameters in this file or the script to experiment with different settings.

Understanding the Results

Results are saved to examples/results/ with the following structure:

Run Metadata

  • run.log - Complete execution log

  • run_configurations.toml - Configuration snapshot for reproducibility

Performance Metrics

  • metrics/ - CSV files and plots showing model performance across windows

Visualizations

  • figures/ - Static plots of fitting results, errors, and comparisons

Animations

  • animation/ - Frame-by-frame visualizations showing model evolution

  • animation/combined_video.gif - Animated overview of the fitting process

Animation Gif
Model Data

  • model_data/ - Fitted parameters per window (CSV format)

  • model_data/*.pkl - Serialized Python objects for later analysis

For detailed output documentation, see Output Format.

Next Steps