The CP-SAT Primer - Using and Understanding Google OR-Tools' CP-SAT Solver
In this section, we will explore various coding patterns that are essential for structuring implementations for optimization problems using CP-SAT. While we will not delve into the modeling of specific problems, our focus will be on demonstrating how to organize your code to enhance its readability and maintainability. These practices are crucial for developing robust and scalable optimization solutions that can be easily understood, modified, and extended by other developers. We will concentrate on basic patterns, as more complex patterns are better understood within the context of larger problems and are beyond the scope of this primer.
The naming conventions for patterns in optimization problems are not standardized. There is no comprehensive guide on coding patterns for optimization issues, and my insights are primarily based on personal experience. Most online examples tend to focus solely on the model, often presented as Jupyter notebooks or sequential scripts. The gurobi-optimods provide the closest examples to production-ready code that I am aware of, yet they offer limited guidance on code structuring. I aim to address this gap, which many find challenging, though it is important to note that my approach is highly opinionated.
For straightforward optimization problems, encapsulating the model creation and solving within a single function is a practical approach. This method is best suited for simpler cases due to its straightforward nature but lacks flexibility for more complex scenarios. Parameters such as the time limit and optimality tolerance can be customized via keyword arguments with default values.
