IFU

The full list of required packages is available in requirements.txt. Run pip install -r requirements.txt or conda install --file requirements.txt to install them all at once.

The code is divided into two main folders:

"Task12": containing two files both combining tasks 1 and 2 of the project: data cleaning, predictive model choice and construction of the three cost matrices (distance, time and financial cost):

• Task12.ipynb which uses 2015 data (thus GPS coordinates) to predict the three possible cost matrices and;
• Task12-2020.ipynb which uses 2020 data instead (thus zones, more generally) to do the same.

"Task3": containing the code for the three routing algorithms and the visualization; the former output the routes in a CSV file, the latter uses it as an input to display the routes on a map:

• routing_optimal.ipynb contains the code for the optimal routing algorithm;
• routing_closest_neighbor.ipynb contains the code for the closest neighbor routing algorithm;
• routing_savings.ipynb contains the code for the savings routing algorithm;
• maps.ipynb contains the code for the visualization of the routes on a map.

The cells can be run sequentially in the order they're presented in each file, starting with generating the cost matrices CSV files in Task12.