jcanny edited this page Mar 9, 2015 · 35 revisions

Table of Contents

BIDMat Matrix Library Documentation

New: API Docs

New: BIDMat Quick Reference


BIDMat is a matrix library+interactive environment intended to support large-scale exploratory data analysis. It includes GPU and cluster computing support and is a sibling of BIDMach, a machine learning Toolkit. Together BIDMat and BIDMach on single nodes hold the performance records (including vs. cluster systems) for a large and growing list of machine learning problems. Here are our current benchmarks. Here is the project home page which includes links to compiled code bundles for both toolkits.

This is the main documentation page. Help topics are listed below.



Matrix Algebra

Complex Matrices

Loading and Saving Matrices

Sparse Matrices

String Matrices


GPU Matrices

Generic Matrices

Matrix Caching

Matrix Functions

Scientific Functions

Linear Solvers and Eigenvalues

Random Number Generators