CBLIB - The Conic Benchmark Library.
This website hosts a collection of benchmark problems for conic mixed-integer and continuous optimization. The Conic Benchmark Library (CBLIB) was initiated in 2012 by Erling D. Andersen and Henrik A. Friberg from MOSEK ApS, and has been hosted ever since at the Department of Optimization at Zuse Institute Berlin. The goal of CBLIB is to encourage algorithmic research, support software improvements regarding reliability and performance, and ultimately become a standard test set for comparative studies, very much in the spirit of the Netlib LP and MIPLIB collections.
How to obtain the library.
The CBLIB 2014 library was constructed with a primary focus on real and realistic optimization problems with second-order cones. It contains 121 instances out of which 80 are mixed-integer and can be used a as standard reference for benchmarking. See this paper for details:
- Henrik A. Friberg. CBLIB 2014: A benchmark library for conic mixed-integer and continuous optimization. Mathematical Programming Computation, 8(2):191-214, 2016. [final, erratum and early draft]
Since this publication, however, many instances have been contributed to CBLIB and our website now contains 1526 instances out of which 1455 are mixed-integer.Downloads:
- CBLIB base of scripts and tools (zipfile, <1MB). [recommended for benchmarking; see details on github]
- The complete list of continuous instances.
- The complete list of mixed-integer instances.
- CBLIB 2014 with continuous (tarball, 4.7 GB) and mixed-integer (tarball, 14 MB) instances.
Call for instances: Please consider contributing. If you have interesting industrial or academic instances, or if you can point us to publicly available, not yet included instances, please contact us. We are happy to help with the process of making instances anonymous if necessary and converting them to suitable formats.
The CBF format (see its technical manual and its description in the CBLIB 2014 paper) has been developed to simplify benchmarking over many different—and often incompatible—solvers. For this reason it has been used throughout CBLIB, and we embrace software that makes use of it:
- The cbftool that allows for simple conversion from this to several other file formats.
- The CBF parsers written in MATLAB, Python and C/C++.
- The PICOS modelling language with CBF read and write capability.
Note: Extensions to the CBF format has been proposed and are backwards compatible. [version 2 proposal]
For questions, hints, and contributions, please contact Henrik A. Friberg or Felipe Serrano.
See tweets by @cblibtw.
Our free and open license policy.
© 2012-2016 by Zuse Institute Berlin. All rights reserved.
Redistribution and use in source and binary forms, with or without modification, are permitted provided that the following conditions are met: 1. Redistributions of source code must retain the above copyright notice, this list of conditions and the following disclaimer. 2. Redistributions in binary form must reproduce the above copyright notice, this list of conditions and the following disclaimer in the documentation and/or other materials provided with the distribution. 3. Neither the names of the copyright holders nor the contributors may be used to endorse or promote products derived from this software without specific prior written permission.
THESE DATA AND SOFTWARE ARE PROVIDED "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT HOLDERS AND CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.