An ensemble of different time series anomaly detection methods is being constructed to estimate the probability of an anomaly at time
├── app.py
├── docker-compose.yaml
├── Dockerfile
├── environment.yml
├── README.md
├── requirements.txt
├── run_app_locally.sh
└── src
├── aac_ts_anomaly
│ ├── config
│ │ ├── aws_config.py
│ │ ├── global_config.py
│ │ ├── __init__.py
│ ├── data
│ ├── resources
│ ├── services
│ │ ├── file_aws.py
│ │ ├── file.py
│ │ ├── __init__.py
│ └── utils
├── __init__.py
├── notebooks
├── setup.py
└── templates
#conda env create -f environment.yml # optionally
conda create -n env_tsanomaly
conda activate env_tsanomaly
To install the package locally, execute the following steps:
cd aac_ts_anomaly
pip install -r requirements.txt # in case no environment.yml was used
pip install -e src
Start streamlit application by running:
bash run_app_locally.sh
Build image and start container:
docker-compose up -d