This is a python implementation of the User Equilibrium and Social Optimum Traffic Assignment using the Frank-Wolfe Algorithm. It uses the python-igraph library to calculate the all_or_nothing assignments in each iterations. It includes implementation for an LA network and Chicago Regional Network. The implementation also allows to increase/decrease the demand by specifying a ratio.

For example to run the program on the LA network:
python Demand_Study.py LA 0.5 UE
The first argument LA specifies the name of the network, 0.5 means that the original demand will be halved, UE indicates that this is User Equilibrium case

As another example:
python Demand_Study.py ChicagoRegional 1 SO
This command runs the Frank-Wolfe algorithm on the Chicago Regional Network, with the original demand (demand X 1), and looking for the Social Optimum SO.

For access to network data for the LA network, please send email to [email protected]
The data for the chicago network is available at this github account