Docker contaienr for optimization with pyomo

This is repository of docker container for optimization with pyomo and some of MINLP solvers.

Solvers

• Bonmin
• CBC
• Couenne
• GLPK
• IpOpt
• SCIP

How to use

Requirement

• Source code of SCIP
• Docker

Check license status

Although this repository is distributed under BSD-3-Clause lisence, some of solvers have restriction to use.

• The solvers distributed on Coin-OR were under Eclipse Public Lisence. You can use them by free.
• As SCIP is distributed under ZIP Academic lisence, you cannot use it by free if it is commercial usage. Then, please consult to official support.

Prepare third party software

This container requires source code of SCIP solver. Please download it and place into the repository. When you want to put out SCIP, please comment out the related procedure in Dockerfile.

Setting library version

The default version of solvers are

• BOMIN: 1.8.7
• COUENE: 0.5.7
• IPOPT: 3.12.12
• SCIP: 6.0.1
• CBC and GLPK: Latest version by apt-get

When you use different version, you have to tell them to docker. Mainly, there are three solutions.

1. Modify docker-compose.yml
2. Pass environment variable when you run
3. Place .env file i nthe directory.

See official document of docker.

Running container

docker-compose up -d

And you can access jupyter lab in your browser!