SmartManPy

SmartManPy focuses on simulating complex features that evolve over time.

This project extends the Python 3 port of ManPy by Data Revenue GmbH (https://github.com/datarevenue-berlin/manpy).

ManPy was originally developed as part of the FP7 DREAM project: https://www.manpy-simulation.org/, which was released under the LGPL license

We performed the following modifications/additions (non-exhaustive):

  • Added complex features to machines:

    • Features can be drawn from a variety of random distributions

    • Simulation of data drift / shifting distributions

    • Failures based on feature values

    • Different random distributions for different entity labels

    • Functional dependencies between features

    • Timeseries that can be interpolated from a couple of real-world data points

  • Quality control based on feature values or distributions

  • Database connection

  • CSV export

License: LGPL

How to get started

Follow the steps below to install SmartManPy:

  1. Clone the Git repository to the desired directory (we call it MANPYDIR in the following)

  2. Open your command line and navigate to MANPYDIR

  3. Activate the virtualenv/conda env of your choice

  4. Enter the following command:

pip install -e .

  1. You can now import manpy:

import manpy

Documentation

Most important modules