smoek 1.0.0

A lightweight optimization modeling library.

Overview

Smoek is an experimental Python modeling environment that uses a compact abstract model expression to enable fast transformation of model structure, deferred model instantiation, and code generation. Smoek was designed to support lightweight model declarations that can be transformed into back-end implementations such as Pyomo, Poek, Coek, and other modeling languages or specialized solver interfaces. Abstract model descriptions can be written compactly in Python, transformed quickly, and used to generate efficient concrete optimization models in lower-level or performance-oriented modeling systems. The overhead for Smoek declarations and problem transformations is minimal, and low- level modeling libraries like Coek can be leveraged to significantly accelerate model generation.

Installation

Basic Installation

Install smoek with pip:

pip install smoek

With Optional Backends

Smoek supports multiple optimization backends. Install with optional dependencies:

# Install with poek backend
pip install smoek[poek]

# Install with coek backend  
pip install smoek[coek]

# Install with all backends
pip install smoek[backends]

Development Installation

For contributors and developers:

git clone https://github.com/sandialabs/smoek.git
cd smoek
conda env create -f dev_environment.yml
conda activate smoek_dev
pip install -e ".[dev]"

See CONTRIBUTING.md for detailed development setup and guidelines.

Requirements

• Python 3.10 or higher
• Pyomo >= 6.0
• munch
• numpy >= 1.20

Testing

Smoek tests can be executed using pytest:

cd smoek
pytest .

If the pytest-cov package is installed, pytest can provide coverage statistics:

cd smoek
pytest --cov=smoek .

The following options list the lines that are missing from coverage tests:

cd smoek
pytest --cov=smoek --cov-report term-missing .

Note that pytest coverage includes coverage of test files themselves. This gives a somewhat skewed sense of coverage for the code base, but it helps identify tests that are omitted or not executed completely.