Appendix B. Mahout math
Many algorithms implemented by Mahout rely heavily on vector and matrix math. Mahout has its own self-contained vector and matrix math library in the math module that can be reused separately from Mahout itself. It is actually an adapted form of CERN’s Colt library (http://dsd.lbl.gov/~hoschek/colt/).
This appendix provides a practical overview of key parts of this math module that users encounter frequently while using Mahout: Vector and Matrix instances, and their associated operations. It is not exhaustive documentation; the Mahout project Javadoc and source code can provide more detail if you’re interested.
Vector means slightly different things in different contexts, though all of them are similar. Most people encounter a vector first in physics, where it denotes a direction and is typically illustrated by an arrow. For our purposes, we sometimes use vectors to represent points instead. This isn’t such a different idea; a point merely corresponds to a vector from the origin to that point.