Anomaly Detection for mHealth Data

Files

mhealth_anomaly_detection/: functions used to simulate and do anomaly detection on mHealth dataset
- anomaly_detection.py: Classes for rolling window based anomaly detection
- datasets.py: Loads and pre-processes external datasets, currently loads the CrossCheck daily dataset
- format_axis.py: Helper function to format plot axes
- load_refs.py: Helper functions to load static files
- plots.py: Functions used to generate figures
- simulate_daily.py: Simulator of daily mHealth data
tests/: Unit tests for functions
cache/: holds simulated anomaly data
lib/: holds static referenced files
- colors.json: Color palettes used for different plots
- feature_parameters.json: Statistical parameters for simulated features
output/: Contains plots and results
run_scripts/: Scripts running and combining functions within mHealth_anomaly_detection folder to generate results
notebooks/: Jupyter Notebooks for example analyses
poetry.lock: Lock file used by poetry for python environment
pyproject.toml: Packages and requirements for project