jMarkov is a Java framework for Markov chain (MC) modeling that provides the user with the ability to define MCs from the basic rules that determine their dynamics. From these rules, jMarkov explores and builds the state space and the other MC parameters, which are then used to solve the MC and determine user-defined steady-state and transient performance measures.
jMarkov has the following modules:
- jMarkov Core Module: Permits modeling large-scale finite Markov Chains
- jQBD: Modeling of Quasi Birth and Death processes (QBDs)
- jPhase: Modeling of Phase type distributions
- jMDP: Modeling of Markov Decision Processes
The source code of jMarkov releases can be downloaded from the following link:
The jMarkov library needs some third party libraries for all its features to work properly. All the libraries listed here are open source, but keep in mind that not all of them are released under the Eclipse Public license, as it is jMarkov. We recommend to verify that the conditions of the third party libraries licenses listed here fit the requirements of your project.
- COLT: http://acs.lbl.gov/ACSSoftware/colt/
- JAMA: http://math.nist.gov/javanumerics/jama/
- JCOMMON: http://www.jfree.org/jcommon/
- JFREECHART: http://www.jfree.org/jfreechart/
- MTJ: https://github.com/fommil/matrix-toolkits-java
- QSOPT: http://www.math.uwaterloo.ca/~bico/qsopt/
- SSJ: http://simul.iro.umontreal.ca/ssj-2/indexe.html.
The following module manuals are available:
jMarkov is maintained by:
- Julio C. Góez (jgoez at jmarkov dot org): Core Module, jQBD
- Juan F. Pérez (jfperez at jmarkov dot org): Core Module, jPhase
- Daniel Silva (dfsilva at jmarkov dot org): jMDP
- Juan P. Alvarado,
- Diego Bello,
- Rodrigo Cáliz,
- Marco Cote,
- Julio C. Góez (Norwegian School of Economics, Bergen, Norway),
- Leonardo Lozano,
- Juan F. Pérez (Universidad del Rosario, Bogotá, Colombia),
- Germán Riaño,
- Andrés Sarmiento Plata,
- Andrés Sarmiento Romero,
- Daniel Silva (Auburn University, Auburn, Alabama, USA),
- Laura Vielma