MultiwayNumberPartitioning

A simple Julia package to optimally solve the multiway number partitioning problem using a JuMP model with mixed-integer programming.

There is one main function partition which tries to accomplish the following task: given a collection of numbers S and a number k, try to partition S into k subsets of roughly equal sum.

For example:

julia> using MultiwayNumberPartitioning, HiGHS

julia> S =  [1, 1, 1, 3, 2, 1];

julia> inds = partition(S, 3; optimizer = HiGHS.Optimizer)
6-element Vector{Int64}:
 2
 2
 3
 1
 3
 2

julia> S[inds .== 1] # group 1
1-element Vector{Int64}:
 3

julia> S[inds .== 2] # group 2
3-element Vector{Int64}:
 1
 1
 1

julia> S[inds .== 3] # group 3
2-element Vector{Int64}:
 1
 2

We can see all three groups here have equal sum.

See the example for a more detailed usage example.

Choice of objective function

We can choose various objective functions for the algorithm to use during the optimization procedure when finding a partitioning configuration.

MultiwayNumberPartitioning.jl provides three objective functions:

partition_min_largest!: minimize sum of the largest subset
partition_max_smallest!: maximize the sum of the smallest subset
partition_min_range!: minimize the difference between the sum of the largest subset and the smallest

(where here "largest" and "smallest" refer to the sums of the subsets).