LILAC is a fairly simple program to run. It takes few command line parameters, instead utilizing input files to control program execution.

It can take two input parameters-the name of the output file, and the index in that order. Both are optional parameters, but one must specify the output file if the index is to be specified.

If no input parameters are specified, the program runs with an output file of data_out.out, and an index of zero. This is equivilant to a call like ./lilac data_out.out 0

If the output file is passed, then the output will be contained in the filename, and if the index is passed, then LILAC will run at that index.

When running at a non-zero index, LILAC does the following:

1. It takes the number of iterations that a single run is supposed to have
2. Finds out how many iterations should happen if a single core were to reach the current run#
3. Performs that many fake iterations

A fake iteration means that the engine moves forwards a step in the parameter space, but doesn't perform a simulation or update any of the internal classes.

For running on distributed platforms, it is often best to run a script instead of directly running LILAC. This script could create a custom filename for each index, and then call LILAC with the proper parameters. Examples can be found in run_dir.