Bringing real and simulated data together


JAS-mine is a Java platform that aims at providing a unique simulation tool for discrete-event simulations, including agent-based and microsimulation models.

With the aim to develop large-scale, data-driven models, the main architectural choice of JAS-mine is to use whenever possible standard, open-source tools already available in the software development community.

The main value added of the platform are: (1) integration of I/O communication tools, in the form of embedded RDBMS (relational database management systems) tools and automatic CSV table creation, (2) advanced multi-run tools to facilitate design of experiments (DOE), (3) sophisticated regression libraries that allow a complete separation of regression specifications from the code, and permit uncertainty analysis of the model outcome by bootstrapping the estimated coefficients across different simulation runs.

JAS-mine allows to separate data representation and management, which is automatically taken care of by the simulation engine, from the implementation of processes and behavioral algorithms, which should be the primary concern of the modeler. This results in quicker, more robust and more transparent model building.

Features

  • A discrete-event simulation engine, allowing for both discrete-time and continuous-time simulation modeling
  • A Model-Collector-Observer structure
  • Interactive (GUI based) batch and multi-run execution modes, the latter allowing for detailed design of experiments (DOE)
  • A library implementing a number of different matching methods, to match different lists of agents
  • A library implementing a number of different alignment methods (including binary and multiple choice alignment), to force the microsimulation outcomes meeting some exogenous aggregate targets
  • A Regression library implementing a number of common econometric models, from continuous response linear regression models to binomial and multinomial logit and probit models, which includes automatic bootstrapping of the coefficients for uncertainty analysis of the model oucomes
  • A statistical package based on the cern.jet libraries
  • Embedded H2 database
  • Export to .CSV files as a faster alternative to database persistence
  • MS Excel I/O communication tools
  • Automatic GUI creation for parameters by using Java annotation
  • Automatic output database creation
  • Automatic agents’ sampling and recording
  • Powerful probes for real-time statistical analysis and data collection
  • A rich graphical library for real-time plotting of simulation outcomes
  • Eclipse plugin, which allows to create a JAS-mine project in just a few clicks, with template classes organised in the JAS-mine standard package and folder structure
  • Maven version control

What to do next