MADlib^® is an open-source library for scalable in-database analytics. It provides data-parallel implementations of mathematical, statistical and machine learning methods for structured and unstructured data.

Installation and Contribution

See the project webpage MADlib Home for links to the latest binary and source packages. For installation and contribution guides, please see MADlib Wiki

User and Developer Documentation

The latest documentation of MADlib modules can be found at MADlib Docs or can be accessed directly from the MADlib installation directory by opening doc/user/html/index.html.

Architecture

The following block-diagram gives a high-level overview of MADlib's architecture.

Third Party Components

MADlib incorporates material from the following third-party components

1. argparse 1.2.1 "provides an easy, declarative interface for creating command line tools"
2. Boost 1.47.0 (or newer) "provides peer-reviewed portable C++ source libraries"
3. doxypy 0.4.2 "is an input filter for Doxygen"
4. Eigen 3.2.2 "is a C++ template library for linear algebra"
5. PyYAML 3.10 "is a YAML parser and emitter for Python"
6. PyXB 1.2.4 "is a Python library for XML Schema Bindings"

Licensing

License information regarding MADlib and included third-party libraries can be found inside the license directory.

Release Notes

Changes between MADlib versions are described in the ReleaseNotes.txt file.

Papers and Talks

• MAD Skills : New Analysis Practices for Big Data (VLDB 2009)
• Hybrid In-Database Inference for Declarative Information Extraction (SIGMOD 2011)
• Towards a Unified Architecture for In-Database Analytics (SIGMOD 2012)
• The MADlib Analytics Library or MAD Skills, the SQL (VLDB 2012)

Related Software

• PivotalR - PivotalR also lets the user run the functions of the open-source big-data machine learning package MADlib directly from R.
• PyMADlib - PyMADlib is a python wrapper for MADlib, which brings you the power and flexibility of python with the number crunching power of MADlib.