Calculate Supply Chain Visibility with Sparse Data

The repository is used to generate the results of the paper: van Schilt, I. M., Kwakkel, J. H., Mense, J. P., & Verbraeck, A. (2024). Dimensions of data sparseness and their effect on supply chain visibility. Computers & Industrial Engineering, 191, 110108. doi:10.1016/j.cie.2024.110108 It present the code on how to calculate supply chain visibility for a given supply chain network with sparse data.

This repository is also part of the Ph.D. thesis of Isabelle M. van Schilt, Delft University of Technology. The version of the code used in the Ph.D. thesis is available at doi: 10.4121/d491bee7-c911-4099-a60f-075327ebea23.v1.

Content

The repository contains the following files:

• data: This folder contains the ground truth data. This data is generated using a stylized supply chain simulation model in pydsol-model.
• degrade_data: This folder contains the .py files to degrade the ground truth data to sparse data. The data is degraded by removing a percentage of the data on the dimensions of bias, noise, and missing values.
• calculate_scv.py: This python file calculates the supply chain visibility for a given supply chain network with sparse data.
• Run_Visualize_SCV_Individual_Dimensions.ipynb: This Jupyter notebook presents the calculation and the visualization of the supply chain visibility when degrading the individual dimensions.
• Run_Visualize_SCV_Scenarios.ipynb: This Jupyter notebook presents the calculation and the visualization of the supply chain visibility when degrading for scenarios (presented in the paper).
• requirements.txt: This file contains the required packages to run the code.