Partitioning of urban networks to optimize postal delivery routes

This is the repository used for our Introduction to network analysis course at UL-FRI.

Deploy of our project is available at: https://delivery-sys-opt.herokuapp.com/

Project report is available at: report.pdf

Structure

The repository contains multiple subfolders:

• cache/ - the cache folder contains cached files for multiple procedures to increase interactivity of the interactive dashboard.
• data/ - the data folder contains all the datasets in .csv and Pajek format.
• util/ - the lib folder contains helper functions and classes. It is additionally broken down to:
  • dash/ - the dashboard folder contains helper functions for the interactive dashboard. Here, we also specify all the tabs and how they function.
  • lib/ - the lib folder contains the general helper methods and classes.
• notebooks/ - the notebooks folder contains all the Jupyter notebooks used for data exploration and initial analyses.
• scripts/ - the scripts folder contains all the Python scripts used for data exploration and initial analyses.
• assets/ - the assets folder contains assets used in the interactive dashboard.

Documentation

In docs you will find a more comprehensive documentation of our project:

• generation of graphs: Graph generation
• running scripts for graph partitioning: Graph partitioning

Environment setup

To setup the environment for this project, follow these instructions:

1. Install Python version 3.8 or higher and Anaconda.

2. Assuming you have git already installed, clone this repository to the desired destination:

   git clone https://github.com/lzontar/Partitioning-of-urban-networks-to-optimize-postal-delivery-routes.git

3. Move to the folder, where you cloned this repository and import the environment from the config file using pip or conda:

   # Anaconda environment
   conda env create --file environment.yml
   
   # pip environment
   pip install -r requirements-local.txt

4. If using Anaconda, activate environment:

   conda activate Projekt

5. To try out our interactive dashboard in root of the project execute:

   python Dashboard.py 
   

   Open your browser at http://127.0.0.1:3010/ to access the interactive dashboard.

6. Explore the repository by yourself or use our documentation for better understanding.