#machine-learning #trading #framework #crypto #financial-data #finance #math

atelier

a Computational Workshop for Market Microstructure Modeling, Synthetic Simulation and Historical Replay

3 releases

0.1.0 Sep 24, 2023
0.0.87 Oct 7, 2024
0.0.86 Sep 17, 2024
0.0.4 Aug 20, 2024
0.0.1 Oct 22, 2022

#318 in Data structures

Download history 79/week @ 2024-07-29 158/week @ 2024-08-19 182/week @ 2024-09-02 313/week @ 2024-09-09 1024/week @ 2024-09-16 31/week @ 2024-09-23 9/week @ 2024-09-30 175/week @ 2024-10-07

1,423 downloads per month

Apache-2.0

27KB
393 lines

atelier

A Computational Workshop for Market Microstructure Modeling, Synthetic Simulation and Historical Replay.

Overview

Use atelier for modeling market microstructure dynamics, it supports an orderbook-based market structure. Whether you need a high quality reproduction of any given market within a defined period of time (market replay), or, to generate what-if scnearios either completely random, or, with model specification (market simulation).

Contribute

Feel free to contribute, just make sure you check the CONTRIBUTING guidelines. Also consider to pick-up the existing issues.


lib.rs:

Atelier

Provides a computational framework that can be used as a workshop to develop Financial Machine Learning models, especifically battle tested to the level of Market microstructure. Features include:

  • Market Simulation: A synthetic simulation of a market, as detailed as the provided configuration.
  • Market Replay: A high-fidelity reproduction of what actually happened, all the way down to the data granularity that is provided.

Dependencies

~1.3–2.1MB
~43K SLoC