uiuc-490-sp23 0.1.3

Programming assignments for UIUC's ECE490: Introduction to Optimization course during Spring 2023

ECE490

Repo to collaborate within for UIUC's ECE490, Spring 2023

Installation

Installation from PyPI

This should be as simple as:

python -m pip install uiuc-490-sp23

Installation from Source

This project uses Poetry Install it via their instructions and run:

poetry install

from this directory (preferably from within a venv).

Executing

Once installed, individual assignments can be ran by:

python -m uiuc_490_sp23.assignment<n>

where <n> is the assignment number. These will have a CLI implemented using argparse; if you're not sure, just run the above with --help to get a full list of commands.

Assignments

Assignment 1: Gradient Descent

This solves a simple, randomly generated quadratic minimization problem using Gradient Descent. It does so using both fixed step size, as well as an adaptive backtracking line search (Armijo's rule).

Assignment 2: Newton's Method and Projected Gradient Descent

This assignment does two things:

• Implements Newton's method to solve a simple 2D cubic polynomial and plots a fractal displaying the resulting solution based on the starting point
• Implements Projected gradient descent and uses it to solve a simple box constrained optimization problem

Assignment 3: Augmented Lagrangian

This assignment implements the augmented Lagrangian to solve an equality constrained quadratic problem.