Reproducibility dataset for a new family of mixed-integer programming model for irregular strip packing based on vertical slices
Description
This dataset introduces a detailed reproducibility protocol to allow the exact replication of the experiments on Mixed-Integer Programming (MIP) model for irregular strip packing reported in our primary work. Likewise, this dataset provides all raw output data from our experiments used in our primary work. Our reproducibility protocol is based on our RAMNEST V1R3 Java software library of MIP models for irregular strip packing. Our software library implements our new family of NFP-CM-VS MIP models for irregular strip packing and the family of state-of-the-art NFP-CM models introduced by Cherri et al. (2016) and Rodrigues et al. (2017). We evaluate all MIP models in 35 small problem instances, 24 metal instances including pieces with holes, and 11 large problem instances, as detailed in the experimental setup of our work. All MIP models are implemented in our Java software library and solved using the Java API of Gurobi 9.5 with its default parameters. Thus, the use of our software requires the installation of an academic or commercial license of Gurobi 9.5 for UBUNTU. This dataset includes the Java source code and pre-compiled binaries of our software library developed with Netbeans 8.2 for UBUNTU. We also provide all MIP models in LP file format, which allows reproducing our experiments without our software by loading them into Gurobi, or any other MIP solver as CPLEX.
Files
Steps to reproduce
This dataset provides a collection of reproducibility resources to allow the replication of the MIP models, experiments, and results reported in our companion article [1]. We refer the reader to the appendixB.pdf file above, which provides a detailed protocol to reproduce our results. References: [1] J.J. Lastra-Díaz, M.T. Ortuño, A new mixed-integer programming model for irregular strip packing based on vertical slices with a reproducible survey, Submitted for Publication. (2022).