Skip to content
nimgould edited this page Nov 9, 2022 · 11 revisions

CUTEst, a Constrained and Unconstrained Testing Environment on steroids

CUTEst is the latest evolution of CUTE, the constrained and unconstrained testing environment for numerical optimization.

Linux and OSX users wishing to install with gcc/gfortran may install CUTEst using the Homebrew package manager.

To use CUTEst, you will need to install three packages, ARCHDefs, SIFDecode and CUTEst. To keep things simple, install all three packages in the same directory:

  • To download ARCHDefs using svn,

             svn checkout --username anonymous https://github.com/ralna/ARCHDefs/trunk ./archdefs

         (with a blank password if prompted), or, using git,

             git clone https://github.com/ralna/ARCHDefs ./archdefs

         or go to the web page 

              https://github.com/optimizers/archdefs-mirror/releases

         and download/unpack one of the available stable zip or tar bundles.

  • To download SIFDecode using svn,

             svn checkout --username anonymous https://github.com/ralna/SIFDecode/trunk ./sifdecode

         (with a blank password if prompted), or, using git,

             git clone https://github.com/ralna/SIFDecode ./sifdecode

         or go to the web page 

             https://github.com/optimizers/sifdecode-mirror/releases

          and download/unpack one of the available stable zip or tar bundles.

  • To download CUTEst using svn,

             svn checkout --username anonymous https://github.com/ralna/CUTEst/trunk ./cutest

         (with a blank password if prompted), or, using git,

             git clone https://github.com/ralna/CUTEst ./cutest

          or go to the web page

             https://github.com/optimizers/cutest-mirror/releases

          and download/unpack one of the available stable zip or tar bundles.

Next, set the environment variables $ARCHDEFS, $SIFDECODE and $CUTEST, respectively, to point to the instalation directories used above. If you intend to use the Matlab interfaces, also set the environment variable $MYMATLAB to  point to the directory containing your installed Matlab; Matlab's mex executable should be found under $MYMATLAB/bin. You may also wish to consult the README in $CUTEST.

Now move to the ./cutest directory, enter the command $ARCHDEFS/install_optrove and follow the instructions.

Remember to set the environment variables as indicated by the installation scripts.

To obtain the test problem set,  either download it using

             git clone https://bitbucket.org/optrove/sif ./sif

or download/unpack one of the available stable zip or tar bundles from 

           https://gitlab.com/dpo/mastsif-mirror/tags

or see

            https://www.cuter.rl.ac.uk/mastsif.html

A further set of quadratic programming test problems by Maros and Meszaros may be downloaded using

             git clone https://bitbucket.org/optrove/maros-meszaros ./marosmeszaros

or download/unpack one of the available stable zip or tar bundles from 

           https://github.com/optimizers/maros-meszaros-mirror/releases

or see

           https://www.cuter.rl.ac.uk/marosmeszaros/marmes.html

A short description of the design and scope of CUTEst is available as

            http://purl.org/net/epubs/work/65540.

If you wish to use CUTEst with python, Jari Fowkes and Lindon Roberts from Oxford University have kindly provided an interface:

            https://github.com/jfowkes/pycutest.

In order to uninstall a particular brand of CUTEst, execute

            $CUTEST/bin/uninstall_cutest

To completely remove CUTEst, execute

            rm -rf $CUTEST ~/.cutest

Clone this wiki locally