---
tags : A0S4, assignment
---
# AOS4 - Assignment #2
## Scalarization techniques for combinatorial multiobjective problems
1. Install e.g. *Python 3.9* and the PuLP library.
2. Decide on a *representation for multi-valuated graphs*.
3. Write a function mapping an input graph and valuation to a *LP formulation of Shortest Path*. Test it on the biobjective graph seen during session #2.
4. Write a function allowing to solve *Shortest Path for a convex combination of the objectives*. Test it. Note there are unsupported efficient solutions.
5. Write a function performing *$\epsilon$-constraint on a bi-objective LP formulation*. Test it.
6. What obstacle (hinted at during the tutorial) arises? Can you find a workaround? Is it efficient? Can you do much better?