Implementation in C++ (requires CUDA 9 or higher), with Python >=3.7 scripts.

Once downloaded the whole project folder, compile the binaries.

On Linux machines:

• make eress
• make kress
• make homgen

The following parameters must be passed as arguments to homgen:

• path/to/inputfile
• --hs_output_file:path/to/hsfile to specify the path to the homogeneous system file to be created

The following parameters must be passed as arguments to eress:

• path/to/inputfile
• --tc or --zi, depending on the index that one wants to use (Tc or zI)
• --n_results:number to specify the number of reported results
• --output_file:path/to/outputfile
• --hs_input_file:path/to/hsfile to specify the path to the file of the homogeneous system (created by means of homgen)

The following paramaters must be defined and passed as arguments to kress:

• --tc or --zi, depending on the index that one wants to use (Tc or zI)
• --dimension:number to specify the search space dimension
• --swarm_size:number to specify the swarm size
• --n_seeds:number to specify the number of seeds
• --range:number to specify the range
• --n_iterations:number to specify the number of iterations
• --kmeans_interv:number to specify the kmeans interval
• --print_interv:number to specify the print interval
• --n_results:number to specify the number of reported results
• --rseed:number to specify the seed for the pseudorandom number generator
• --input_file:path/to/inputfile
• --output_file:path/to/outputfile
• --hs_input_file:path/to/hsfile to specify the path to the file of the homogeneous system (created by means of homgen)
• --comp_on:0/1

To run the tools, edit and use the ress.py script (in the scripts/ folder):

• python ress.py