Configuration

Various configuration options are handled using environment variables. These can easily be configured using .env files. You can check out python-dotenv for more information on how to use .env files with Python.

How to? For the basics, create a file called .env in the root of your project and add the following:

A simple .env file.

IESOPT_CORE = 2.0.0

Tip

For many applications it is perfectly fine not to use any configuration at all. The defaults are chosen to work well in most cases. Each release of the Python wrapper iesopt is bound to a specific version of the Julia package IESopt.jl. Changing this is only necessary if you want to use a different version of the Julia package. Further, we install a default version of HiGHS as solver - so unless you need to use (and have access to) a commercial solver, you should be fine with the default settings.

Julia packages

IESopt.jl

To install a specific version of the Julia core, IESopt.jl, use the following:

Setting the IESopt.jl version.

IESOPT_CORE = 2.0.0

Solvers

Installing solver packages can be done using the following:

Setting the solver versions.

IESOPT_SOLVER_HIGHS = 1.12.0

Other examples may be setting IESOPT_SOLVER_CPLEX or IESOPT_SOLVER_GUROBI.

Important versions

The following other entries are potentially used in some projects:

Fine grained version control.

IESOPT_JULIA = 1.11.1
IESOPT_JUMP = 1.23.3

Arbitrary packages

To install packages that are not part of or related to IESopt, you can use the following:

Arbitrary packages.

IESOPT_PKG_ModelingToolkit = 1.0.3

Make sure the proper casing of the package name is used. The package name is case-sensitive. Even though environment variables are commonly upper-case only, the package name has to reflect the Julia package’s wording.

Non-registered packages

To install non-registered packages, you can use the following syntax:

Installing packages from GitHub.

IESOPT_CORE = https://github.com/ait-energy/IESopt.jl#super-important-feature

When such a repository is updated, we might not be able to automatically update the package. In this case, you can use the following

Updating packages from GitHub.

import iesopt
iesopt.julia.seval('import Pkg; Pkg.update("IESopt")')

to forcefully update a Julia package.

Options

Currently the following options are available:

Options:

IESOPT_MULTITHREADED:

yes or no (default). Talk to us before using this.

IESOPT_OPTIMIZATION:

rapid, latency (default), normal, or performance. Consider using latency for small models, or repeatedly executing your code, since it may be faster for these kind of work loads. Set it to performance for large models where initial up-front costs are not relevant. normal refers to the default settings chosen by “just launching Julia”. For iterative development on a single small model, rapid might be the best choice - however, it compromises actual performance, even in subsequent runs, so it is not recommended for production code.