QUBO Instances for benchmarking
This package provides a wrapper around QUBOLibData.jl's artifact to easily access QUBO / Ising instances. Instances are retrieved as model objects from QUBOTools.jl.
julia> import Pkg; Pkg.add(url="https://github.com/pedromxavier/QUBOLib.jl")
julia> using QUBOLib
julia> coll = first(list_collections()) # get code of the first registered collection
julia> inst = first(list_instances(coll)) # get code of the first instance from that collection
julia> model = load_instance(coll, inst) # retrieve QUBOTools model
Warning This requires SQLite.jl and DataFrames.jl to be installed.
julia> using SQLite, DataFrames
julia> db = QUBOLib.database()
julia> df = DBInterface.execute(
db,
"SELECT collection, instance FROM instances WHERE size BETWEEN 100 AND 200;"
) |> DataFrame
julia> models = [
load_instance(coll, inst)
for (coll, inst) in zip(df[!, :collection], df[!, :instance])
]