If the GUI version is being run, the output is displayed as it is produced in the window. As described above, there are various options to control this.
For both GUI and non-GUI versions, an output file is generated, along with a file containing gnuplot commands. The gnuplot program is available separately, and is available from www.gnuplot.org . To view the output with gnuplot, in a terminal window issue the command gnuplot xxx.gpi, where xxx is the base name of the input and output files. The gnuplot command file is generated before the simulation begins, so that users may begin viewing the data as it is generated. This means however that the run length specified initially is used by gnuplot to scale the time axis, and consequently the initial data may appear to be quite compressed on the left hand side of the plot. This may easily be altered by editing the gnuplot command file (or a copy) to reflect the time range the user wishes to view, and running gnuplot with the edited version. The user will probably also wish to do this if the run ends before the initially specified time is reached.
The format of the output files (.sdo extension) is as follows: The first few lines are preceded by a ``#'', which comments them out as far as gnuplot is concerned. These lines give the number of reactants, and the names of each reactant. Each subsequent line is a line of data from the simulation, beginning with the initial conditions. The first item in each line gives the time, the second the number of steps so far, and the next items the populations of each of the reactants (in the order in which they are named above).
The output file may also be post-processed with the utility ``stodeoutproc''. This prepends three (gnuplot commented, as above) lines to the output, reporting the total time taken by the simulation, and the minimum and maximum time differences between pairs of adjacent steps.
A further utility ``dilute'' takes a completed output file, and produces a second version with fewer data points. The dilution factor is specifiable. This is useful, for example if you have an output file which is very big and oversampled. (Gnuplot doesn't like files which are too big). Bear in mind that thinning a file out by too much can cause important information to be lost. If you have information which may be lost in this way, consider viewing only a small time-slice of the data at a time.