This page hosts the documentation of the SmartCGMS software architecture and all of its components.


The console application lets you to execute a configuration, or to optimize filter parameters. Note that some filters can modify the configuration at runtime. For example, Pattern-Prediction can learn during execution, while saving the learned patterns to the configuration. To save what it has learned, use the /save_configuration switch.

     Usage: console3 filename [/optimize=filter_index,parameters_name[ /solver=generations[,population[,solver_GUID]]]] [/save_configuration]

     The filename designates the configuration of an experimental setup. Usually, it is .ini file.

     The filter_index starts at zero. parameters_name is a string.

     Generations is the maximum number of generations to evolve/computational steps.
     Population is the population size, when applicable, or 0 to allow entering the solver_id.
     Solver_GUID is solver's id, formatted like {01274B08-F721-42BC-A562-0556714C5685}.
     Make no spaces around the , delimiters.
     Default solver is Halton-driven Meta-Differential Evolution, 1000 generations, 100 population size.

     Available solvers:
     {1B21B62F-7C6C-4027-89BC-687D8BD32B3C} - MT MetaDE
     {01274B08-F721-42BC-A562-0556714C5685} - Halton MetaDE
     {2332F9A7-39A2-4FD6-90D5-90B885201869} - RND MetaDE
     {93FFF43A-50E8-4C7B-82F0-F290DFF2089C} - Halton PSO
     {033D92B0-B49C-45D1-957F-57682D56ABD2} - Sequential Brute Force Scan
     {FA42286B-928C-47BD-AAF3-2FE47377466D} - Sequential Convex Scan
     {6BAD021F-6F68-4246-A4E6-7B1950CA71CB} - RumorOpt

More solvers can be available, depending on your configruation and library-availability. SmartCGMS can use the NLOpt and PaGMO libraries.