Skip to content

Latest commit

 

History

History
77 lines (57 loc) · 3.02 KB

File metadata and controls

77 lines (57 loc) · 3.02 KB

optimisation-models

Collection of optimisation test models (LP/MIP/QP) packaged as classpath resources. Used by ojAlgo and related projects for regression testing and benchmarking.

Models with more than 10k variables or constraints, or individual files of 100MB or more, are excluded from all collections to keep the repository and Maven artifact manageable.

Collections

Collection Models Format Problem Types Source
burkardt 7 MPS LP people.sc.fsu.edu/~jburkardt/datasets/mps/mps.html
marosmeszaros 113 SIF Convex QP ftp.numerical.rl.ac.uk/pub/cuter/marosmeszaros
meszaros 71 MPS LP Meszaros LP test set
netlib 97 SIF LP numerical.rl.ac.uk/cute/netlib.html
MIPLIB 495 MPS MIP miplib.zib.de
QPLIB 382 LP + SOL QP / MIQP qplib.zib.de

burkardt

Small LP instances from John Burkardt's MPS dataset collection.

marosmeszaros

The Maros-Meszaros convex QP test set in SIF format. 113 instances included (25 larger ones excluded). All problems have convex quadratic objectives and linear constraints. See marosmeszaros/README.md for per-instance details.

meszaros

The Meszaros LP test set. 71 linear programming instances in MPS format.

netlib

The classic NETLIB LP test set in SIF format. 97 instances included (17 larger ones excluded). Widely used for benchmarking LP solvers. See netlib/README.md for per-instance details.

MIPLIB

Combined collection from all MIPLIB editions (2.0, 3.0, 2003, 2010, 2017), filtered to easy instances with at most 10k rows and 10k columns. 495 mixed-integer programming instances in MPS format. See MIPLIB/README.md for per-instance details and edition membership.

QPLIB

382 quadratic programming instances from QPLIB in LP format with solution files. Covers linear, convex, and non-convex objectives with continuous, binary, integer, and mixed variables. Instances with more than 10k variables or constraints are excluded (71 total), as is one smaller instance whose dense quadratic objective produces a file exceeding 100MB. See QPLIB/README.md for per-instance details, problem type codes, and the full list including excluded instances available from qplib.zib.de.

Usage

Add as a test dependency:

<dependency>
    <groupId>org.ojalgo</groupId>
    <artifactId>optimisation-models</artifactId>
    <version>1.1.0-SNAPSHOT</version>
    <scope>test</scope>
</dependency>

Models are available on the classpath under optimisation/{collection}/.