Project Hadron
Project Hadron is an open-source application framework for in-memory preprocessing and trained model predict, where data analysis, feature optimisation and supervised machine learning, require efficiency and speed within an ML lifecycle. With Apache Arrow as its canonical, and a more directed use of Pandas, Project Hadron offers effective data management, extensive interoperability, improved memory management and hardware optimization.
Within this, Project Hadron looks to improve the availability of objective relevant data, increase the transparency and traceability of data lineage and facilitate knowledge transfer, retrieval and reuse.
For a more in-depth view of Project Hadron read the section on Why Project Hadron? or jump straight to Getting Started once you have installed the package.
Contents:
- Glossary of Terms
- Why Project Hadron?
- What are we solving?
- Where does Project Hadron fit?
- What are its design methodologies?
- Where does it sit within a system pipeline?
- What are capabilities?
- How are capabilities reusable?
- What is a capability pipeline?
- Where can Project Hadron be applied?
- Who would use Project Hadron?
- What is PyArrow?
- Quick glance features
- Installation
- Getting Started
- Introducing Capabilities
- Implementation in Docker
- Capability Intent Actions
- Capability Reports
- Introducing Connectors
- Contributor’s Guide