ISIS

Integrative Systems Implementation Software

What Is ISIS?   Features   Formal Interface   Availability   Authors   References  

1 - What Is ISIS?

ISIS allows you to implement complex, dynamic systems of any structure in a hierarchical manner. It is based on the ModelWorks software (Fischlin 1991 , Fischlin et al. 1994 ) and allows for systems that can consist of any number of models of any type. For instance you can build a system which consists of several coupled ordinary differential equations (DESS - Differential Equation System Specification), difference equations (SQM - Sequential Machines), or continuous-time discrete event submodels (DEVS - Discrete Event System Specification).

ISIS supports the modeling of structured systems in a much more efficient manner than this is possible with ModelWorks:

Firstly, ISIS allows for input/output linking of modular subsystems forming larger supersystems. The subsystems can be specified independently of each other, i.e. there is no need for cross-imports of variables or model objects at the source code level. Subsystems and their objects are uniquely identified by means of identifiers (e.g., modelIdent.stateVariableIdent) and are declared to the ISIS model objects data base. ISIS then links all subsystems at run time by matching all currently declared input variables with the available output variables. The used linking technique is efficient and supports safe and correct data exchange between subsystems. Moreover, it allows for replacing whole subsystems at run time with minimum programing effort, e.g. for structural sensitivity analyses.

Secondly, ISIS allows to fully separate model data (such as the initial values of state variables, parameter values, monitoring attributes etc.) and system structure. The system structure is specified in a Modula-2 program, whereas the model data are stored separately in text files in the form of so-called data frames. These are human readable tables with a flexible, formally defined syntax. The data are automatically retrieved from the data frames and assigned to the models and their objects at run time. The data currently used in the simulation environment can be saved again to disk and stored in a data frame file for documentation or later use in a simulation experiment. ISIS is capable of extracting from a large collection of data frames precisely all data needed by the currently declared (sub-)systems. This makes it possible to flexibly use the same data frames in varying contexts, e.g. when working with different model versions that have similar data needs.

2 - Features

• Modular formulation of model systems supporting hierarchical modeling.
• Separation of model data and system structure; import/export and assignment of model data at run time.
• Arbitrary number of simultaneously integrated continuous, discrete time, discrete event and mixed models of any order limited only by the available memory and the modular formulation of the model systems.
• Interactive simulation environment with graphical user interface. Support of windows, menu technique, buttons and pointing device (mouse) provided by the underlying ModelWorks software.
• ISIS programs are open to extensions of the user- and man-machine interface provided by the library modules of the underlying Dialog Machine or your own routines.

3 - Formal Interface

To learn about the formal interface, notably all functions of ISIS see:

• Overview of the ISIS modules.
• Quick Reference containing all functions of the ISIS in a compact format.

4 - Availability

ISIS is freeware (courtesy ETH Zurich).

It comes as a part of the RAMSES modeling and simulation environment.

Download it for:

• Mac Classic   (part of the RAMSES package)
• Mac OS X   (part of the RASS-OSX package)
• Sun Solaris   (part of the RASS-Sun package)

For information on the needed Modula-2 development environments for the various supported computer platforms see here.

5 - Authors

ISIS has been designed and implemented by Andreas Fischlin with support by Dimitrios Gyalistras, both at ETH Zurich, Switzerland.

6 - Cited References

Fischlin, A. (1991). Interactive Modeling and Simulation of Environmental Systems on Working Stations. In: Möller, D.P.F. & Richter, O. (eds.), Analysis of dynamic systems in medicine, biology, and ecology. Informatik-Fachberichte 275: 131-145. doi: 10.1007/978-3-642-77020-3_17

Fischlin, A., Gyalistras, D., Roth, O., Ulrich, M., Thöny, J. & Nemecek, T., Bugmann, H. & Thommen, F. (1994). Model Works - an interactive simulation environment for personal computers and workstations. Systems Ecology Report No. 14, ETH Zurich, Switzerland, 324pp.

See also the Systems Ecology Publications and Reports.

Modula-2 website ring
List all    |    <<  Prev    |    Next  >>    |    Join