A Graph-Based, Distributed Memory, Modeling Abstraction for Optimization: Data and Code

This dataset contains all code, input and results data relevant to the working paper ["A Graph-Based, Distributed Memory, Modeling Abstraction for Optimization"](https://arxiv.org/abs/2511.14966). This work introduces a general modeling abstraction called a RemoteOptiGraph that is implemented in the open-source package Plasmo.jl. This dataset contains code for running a large-scale capacity expansion model which is solved with Benders decomposition to highlight the operation of this new abstraction. The dataset contains a single zip folder:DistributedPlasmoExamples.zip which contains i) a Julia environment with the packages and their versions used in this study; ii) source code used from an adapted version of GenX.jl; iii) a tutorial example showing how the RemoteOptiGraph can be used; iv) data and code for running a case study corresponding to the western US and solved with Benders decomposition. This case study is modeled with the RemoteOptiGraph abstraction and solves more than 7x faster than solving the monolithic problem without decomposition or the distributed modeling. v) details for working with the RemoteOptiGraph on more complex architectures, such as multiple compute nodes or cloud computing.