What is this?
OR-Tools is Google's open-source software suite for combinatorial optimization: solvers and algorithms for problems like scheduling, routing, packing, and resource allocation.
OR-Tools is Google's open-source software suite for combinatorial optimization: solvers and algorithms for problems like scheduling, routing, packing, and resource allocation.What kinds of problems can it actually solve?Constraint programming (CP and CP-SAT), linear programming (Glop and PDLP), wrappers for mixed integer solvers, bin packing and knapsack, Traveling Salesman and vehicle routing, plus graph algorithms like shortest paths and max flow.
Constraint programming (CP and CP-SAT), linear programming (Glop and PDLP), wrappers for mixed integer solvers, bin packing and knapsack, Traveling Salesman and vehicle routing, plus graph algorithms like shortest paths and max flow.How do I get it installed?Grab a binary release or build from source — the suite supports Make (legacy), CMake, and Bazel builds. After that, the tutorials in the developer guide are the intended entry point.
Grab a binary release or build from source — the suite supports Make (legacy), CMake, and Bazel builds. After that, the tutorials in the developer guide are the intended entry point.I work in Python — do I have to write C++ to use it?No. OR-Tools itself is written in C++, but it ships official wrappers for Python, C# and Java, so you can call the same solvers from those languages.
No. OR-Tools itself is written in C++, but it ships official wrappers for Python, C# and Java, so you can call the same solvers from those languages.Which platforms does it run on?It has been tested on 64-bit Ubuntu 18.04 LTS and up, macOS Mojave with Xcode 9.x, and Windows with Visual Studio 2022, so Linux, macOS, and Windows are all covered.
It has been tested on 64-bit Ubuntu 18.04 LTS and up, macOS Mojave with Xcode 9.x, and Windows with Visual Studio 2022, so Linux, macOS, and Windows are all covered.Do I have to compile it myself, or are there prebuilt binaries?Both options exist — the installation docs cover source builds with Make, CMake or Bazel as well as binary distributions, so you can start with a prebuilt package and skip the toolchain setup.
Both options exist — the installation docs cover source builds with Make, CMake or Bazel as well as binary distributions, so you can start with a prebuilt package and skip the toolchain setup.What is CP-SAT?One of the suite's two constraint programming solvers — it approaches CP problems via Boolean satisfiability. The other is the classic CP constraint solver, which also serves as the routing solver for vehicle routing problems.
One of the suite's two constraint programming solvers — it approaches CP problems via Boolean satisfiability. The other is the classic CP constraint solver, which also serves as the routing solver for vehicle routing problems.Which graph algorithms are included?Shortest paths, min cost flow, max flow, and linear sum assignment, all implemented in the graph component with their own samples.
Shortest paths, min cost flow, max flow, and linear sum assignment, all implemented in the graph component with their own samples.What's the best way to learn it after installing?Follow the tutorials in the developer guide, then read the examples directory — it contains C++, .NET, Java, and Python examples plus Jupyter notebooks.
Follow the tutorials in the developer guide, then read the examples directory — it contains C++, .NET, Java, and Python examples plus Jupyter notebooks.What are similar projects?scipy, networkx, scikit-learn, rebalancer, xgboost, xquad, jax, closure-compiler
scipy, networkx, scikit-learn, rebalancer, xgboost, xquad, jax, closure-compilerWhich topics is it in?scientific-computing, graph, optimization, algorithms, google, c#, bsd, c++, java, apache-2.0, library, python
scientific-computing, graph, optimization, algorithms, google, c#, bsd, c++, java, apache-2.0, library, python