A project that intends to capture arbitrary data sets, evaluate them against different detectors, select the most appropriate and evaluate when new observations are anomalous.
This project is currently under development and taking its very first steps. Features will be documented in the future.
-
Clone this repository.
git clone https://github.com/makingsensetraining/anomaly-detector.git cd anomaly-detector -
Create a conda environment with the dependencies already provided for you.
conda env create -f anomaly-detector.yml
However, if you already had the environment and you just need to update the dependencies, run:
conda env update -f anomaly-detector.yml
-
Activate the environment and get hacking!
activate anomaly-detector
In order to run unit tests, execute in the command line:
runTests