Manage Massive Matrices with Shared Memory and Memory-Mapped Files

Create, store, access, and manipulate massive matrices. Matrices are allocated to shared memory and may use memory-mapped files. Packages 'biganalytics', 'bigtabulate', 'synchronicity', and 'bigalgebra' provide advanced functionality.


News

NEWS for R package bigmemory

See also the Bigmemory Project:

http://www.bigmemory.org

Jay and Mike [email protected]

Jay Emerson [email protected]

Mike Kane [email protected]

Charles Determan [email protected]

2017-11-19 Michael Kane

  • Re-added functions to assign based on column and row names.

2017-11-08 Michael Kane

  • Memory leak problem solved.

2016-09-21 Adam Ryczkowski [email protected]

  • Fixed a bug with handling raw bytes. Now it is possible to use big.matrix to store arbitrary R object (using serialization). If you want to make sure that bigmemory big.matrix handles raw data, add "bigmemory (>=4.5.21)" to the Depends stanza in the Description.

2014-04-15 Mike and Charles [email protected]

  • Enabled use of Rcpp package to simply code and create smaller barrier to adding additional extensions.
  • Added new method to big.matrix objects, allocation_size, to easily check the size of the object in memory.
  • Float type matrices have now been added to enable single precision analytics.

2013-03-09 Jay and Mike [email protected]

  • We missed a NEWS entry. Windows support is temporarily suspended due to issues with the Boost headers.
  • We now depend on bigmemory.sri (a trivial dependency relating to our big.matrix descriptors) and new package BH (for Boost headers).

2011-06-29 Jay and Mike [email protected]

  • Modified deepcopy() with C code (thanks to Zarrar for this)
  • Added -fpermissive to possibly address a compilation issue in Linux
  • Added -lm to address a linking problem with math.h
  • Fixed a problem with reading all 0's in a column in read.big.matrix()
  • We note that there may be an inconsistency with read.big.matrix() and write.big.matrix() that needs to be addressed.

2010-10-30 Jay [email protected]

  • Modified configure to handle a Solaris problem.
  • Added definition of C99FEATURES to isna.hpp; hopefully won't break anything.

2010-05-16 Jay [email protected]

  • Set proper default value op='AND' for mwhich()

2010-05-08 Jay [email protected]

  • The configuration process has been streamlined to avoid warnings and allow for support of Sun Studio compiler on Linux.

2010-05-06 Jay and Mike [email protected]

  • bigmemory >= 2.2.0 is accompanied by sister packages biganalytics and bigtabulate; mutexes (locking) are abstracted to package synchronicity, available now on R-Forge and soon to be uploaded to CRAN. Package bigalgebra is on R-Forge as a beta version while we sort through the range of library configuration options.

Reference manual

It appears you don't have a PDF plugin for this browser. You can click here to download the reference manual.