javaggregation

What is it

This program is a java implementation of algorithms presented in the paper Interpolation by lattice polynomial functions: a polynomial time algorithm, written by myself, Miguel Couceiro and José R. Figueira. It implements distributive lattices, lattice polynomial functions, and the interpolation algorithm presented in the paper. I mainly share these implementations for the sake of transparency ; however, feel free to contact me if you think this code can be of any use to you.

The jar file

It contains a compiled version of the programm, that you can download and launch. The command line for this is

java -jar latticepolynomials.jar [task] [options]

Tasks

The task to execute is specified by the first argument:

• help: displays arguments and option names
• interpolation_test: randomly generates a distributive lattice, a lattice polynomial, and data. From the data, computes the constraints for interpolating lattice polynomials, and display the time spent on this last task. For more details, see Interpolation by lattice polynomial functions: a polynomial time algorithm. Note that this task is executed for each combination of the options -n, -m, -k, -p, -g, -nbrep.
• lattice_gen: Generates random distributive lattices and displays them. One lattice is generated for each combination of the options -k,-p.

Options

• -n arities, default: 1, 2, 3, 4, 5, 6, 10, 15, 20, 30
• -m number of instances in the data, default: 10, 100, 1000, 10000, 50000
• -k height of the lattice, default: 1, 2, 3, 5, 7, 10, 15, 20, 30
• -p rho parameter for lattice generation, default: 0.6
• -g maximal number of focal set of the lattice polynomial, default: 100
• -nbrep, default: 1

Example

java -jar latticepolynomials.jar lattice_gen -k 1,2,3,4,5,7,10 -p 0.5