May 22, 2019

Learn Python, PuLP, Jupyter Notebooks, and Network Design

Some Helpful Exercises From the Domain of Network Design

As part of our book, Supply Chain Network Design, we’ve created a series of mini-case studies. These exercises help you gain intuition on how network design models work.

We’re happy to report that we have now migrated all these models into Python.  If you teach this material, it allows for an easy integration into your class.  Or, if you are reading the book and want to practice the concepts, the Python models make that easy too. Although there are great commercial tools for network design, we decided to use entirely open-source tools so that any student has free and easy access to these cases.  Each is designed so that someone who doesn’t know Python can navigate them with ease.

We realized that there are a lot of supply chain professionals who might already know network design, but are new to Python and want to see what it’s all about.  Likewise, there are a lot of people who have experience with Python, but haven’t explored network design and linear programming.  Even if you’re well-versed in both, these cases can serve as a great refresher on uniting the two in practical ways.

Watson Says... Even if you’re well-versed in both (network design and linear programming), these cases can serve as a great refresher on uniting the two in practical ways. What do you say? Click here to send us your comments

The models and instructions for Python are on this page.  You can find the installation directions in the fifth paragraph.

The setup guide will walk you through the installation of Python (the programming language), Jupyter Notebooks (the interface you’ll use to interact with the models), PuLP (an open-source Python package for optimization problems), and plotly (a Python visualization package that’s used to draw maps).

Once you’ve opened the Jupyter Notebook to access the code, scroll down to the modeling section that uses PuLP.  This serves as a great introduction to formulating a linear (or integer) program in Python.

After the installation is complete, begin with the Chapter 3 exercise on Al’s Athletics.  In this model, you’ll locate warehouses to minimize their average weighted distances from Al’s stores.  Play around with the number of warehouses to see the effect on overall average distance.

The next model I’d suggest investigating  is in the Chapter 6 exercise on UPS, a small parcel shipping model.  In this example, the company has one warehouse in Louisville, and ships five-pound packages around the country. 

Experiment with additional warehouses to minimize the cost - it’s a fun one to think about (for example, why might there be diminishing returns?).

Finally, the Chapter 9 JADE exercise involves a manufacturing model with inbound and outbound costs, different products, and different plants.  This model does a great job giving you a taste of full-scale network design.

By going through these exercises, you’ll learn about Python programming, Jupyter notebooks, PuLP, and network design - you might even have some fun along the way.

(If you are completely new to Python, I suggest checking out Learn Python the Hard Way - after you’re done, come back to the other book exercises and try more!)

Any reaction to this Expert Insight column? Send below.

Your Comments/Feedback