Tutorial
1. Usual Steps of the Usage of this Software
For the Markov reward model analysis a model has to be provided first.
This can be done by opening an existing model through the
Open submenu
in the File menu, or by
creating a new one though the
New submenu in the File menu (alternatively, the
corresponding buttons on the toolbar can be used as well).
Once there is an open model, its analysis window can be opened by
clicking to the Analyse menu,
then
Analyse
Model, and the name of our model. Using this analysis
window different data series can be created. For example, the moments
of the accumulated reward or completion time can be computed as a
function
of time, or as a function of any model parameter. Furthermore, in
case of some Markov reward model types the distribution of these
quantities is also available to compute.
MRMSolve includes a diagram editor, too (Analyse menu,
Open
Diagram Editor submenu), which is used to compare and visualize
the generated data series, even of possibly different models. The
diagram editor is capable of export the plots to .eps format (for
inclusion to a paper, for example), or to gnuplot format (for further
processing).
2. Defining a New Model
In MRMSolve users have two different ways to define the model. With the
so called 'raw description' the
matrices and vectors describing the model have to be given in a direct
manner. For example, all the transitions of the background process and
all the elements of the reward rate vector need to be given. For
lagrer models
there is a very compact and comfortable way of description available,
the so called 'formalizm'.
Using this method, the matrix and vector parameters of the model can be
defined using rules. If the model is well structured, a few rules are
enough to describe the model.
2.1 Raw Model Description
2.2 Using the Formalizm