----------------
Installation
----------------
Download Snoopy's setup file: snoopy-beta-windows-intel-2019-04-26.msi, 
available at Snoopy's website, download section:
https://www-dssz.informatik.tu-cottbus.de/DSSZ/Software/Snoopy#downloads

A double click on the downloaded installer file (msi) should be enough.
The download section also provides detailed installation instructions.

Then, a double click on the Snoopy icon should start Snoopy.

Next, check the Snoopy version you installed:
  -> Help -> About

It should be:
version: 1.30, 
release: beta
built:   Apr 26 2019

For a tutorial video demonstrating FPN modelling and simulation in Snoopy, please watch the video clip:
https://www-dssz.informatik.tu-cottbus.de/video/Video_FPN_24_4_19.mov
also available via 
https://www-dssz.informatik.tu-cottbus.de/DSSZ/Software/Snoopy#manuals

----------------
Typical use case
----------------

The workflow for modelling and simulation is the same for all FPN classes (FCPN/FSPN/FHPN), here explained for Fuzzy Continuous Petri Nets (FCPN); please also compare the corresponding figures in our tool paper.

One example of this type is called Decay Dimerization model which consists of four transitions (reactions) r1-r4 and three places (species) s1-s3. r3 and r4 have the fuzzy kinetic parameters k3 and k4, while r1 and r2 have crisp kinetic parameters. The kinetics of all four transitions follow the mass/action pattern.

This file can be downloaded from:
https://www-dssz.informatik.tu-cottbus.de/DSSZ/Software/Examples
subfolder 'fpn'.

1. Modelling 

The FCPN modelling can be done in one of the following three ways:

1.1 Create a new net by clicking on the 'open a new document' button or
     -> File -> New 
and choose the appropriate net class: 'Fuzzy Continuous Petri Net'.

Then draw the network by dragging the elements of the net (places, transitions) to the canvas and using arcs to connect places and transitions.  A double click on a graph element (place, transition, edge) opens an attribute window, which allows to edit the element-specific properties to specify initial marking of places, kinetic rates (kinetic parameters) of transitions, and other properties.

1.2 Convert the Petri net of a related net class by Snoopy's export feature; an FCPN can be obtained by export from either a CPN, FSPN or FHPN; compare the export relation shown in Fig. 1 of our tool paper.

1.3 One can also load a previously saved FCPN file (file extension '.fcpn'), e.g. Decay_Dimerization.fcpn by clicking on the 'load a document' button, or
    -> File -> Open -> <select your file>

1.4 Constant definition: One can define new constants to specify:
- initial marking,
- kinetic rates (fuzzy number, crisp value).

This can be done by using the Declarations -> Constant Window to specify for each constant:
- name, 
- type (TFN or double), 
- textual description (optional), 
- value (which is a triple for TFNs).

Note: TFNs are only permitted as kinetic parameters.

2. Simulation 

2.1 Open the FCPN simulation window by clicking on View menu and then choosing Start Simulation-Mode: 
     -> View -> Start Simulation-Mode

2.2 One can change the default simulation parameters in the Simulator Configuration subwindow:
- simulation time: interval start, interval end, splitting,
- simulator semantics,
- ODE solver,
- Fuzzy Setting: alpha levels, discretization points and sampling strategy.

Note: Unexperienced users should keep the default values.

2.3. Then, click the 'Start Simulation' button.

2.4 When the simulation finished, double click on the 'default' view to see the simulation results as a fuzzy band for the places selected in the subwindow on the right. Moreover, one can move the scroll bar to view the membership functions of the selected place(s) over time.

2.5 Export results to CSV file: choose CSV Export from the export options in the result viewer window and then click on export button.


------------
Known issues
------------

Due to the fact that the FPN simulation may take a lot of time, we intend to support the Simulation Aborting Feature. This feature allows the user to abort a running simulation. This is useful especially when the user wants to change some simulation parameters e.g, alpha levels without waiting for the whole simulation to be done.

To implement this, we run the fuzzy simulation on a separate thread. Technically speaking, aborting the simulation means terminating the running thread before it has finished. However, while under Windows OS this feature is perfectly working, terminating running threads caused some issues related to memory allocation under MAC OS and Linux.

Thus, FPN in Snoopy is currently working only under Windows OS and will be soon available under MAC OS and some Linux distributions.

-- BTU, Cottbus, 24/04/2019 --