Tools
What you need before the models: the theory that lets you interrogate a solution (duality, shadow prices, reduced costs) and the solver you build and solve the models with.
-
Theory: linear programming
Primal-dual pair in general form, duality rules, weak and strong duality, complementary slackness, shadow prices, sensitivity analysis and the sensitivity protocol used in every chapter.
-
Solver: linear models
The five steps to build a model with
gurobipy, how to run it, how to read the solution, shadow prices, reduced costs and validity ranges.
Linear models and prerequisites
All the models of the course are LP (Linear Programming): linear objective and constraints. An LP is always convex: every local optimum is global and duality holds without exceptions. The prerequisite is the module Linear Programming, in particular duality and sensitivity analysis. Models with a quadratic or nonlinear objective are in the module NLP Modelling.
Notation used throughout the course. Scalars and indices in lowercase (\(x_{it}\), \(\lambda\)); the objects of the models (products, channels, assets, scenarios…) are numbered and the indices run over explicitly enumerated sets, \(i \in \{1, 2, \dots, n\}\); integer counts (\(n \in \mathbb{Z}_{\ge 1}\)), rational data (\(\mathbb{Q}\)); vectors in lowercase bold (\(\boldsymbol{x}\)), matrices in uppercase bold (\(\boldsymbol{A}\)). Dual variables \(\pi_i\), reduced costs \(\bar c_j\), slacks \(\bar s_i\): the bar denotes the values of a feasible solution, the tilde those of an optimal one (\(\tilde x_j\), \(\tilde z\)). In the models the wording is always "subject to", the variables are introduced before the formulation and the constraints defining them close the model.