Special Topics - Computational Methods in Physics (PHYS*7730)

Course Information

Instructor

Alex Gezerlis (MacN 219, gezerlis@uoguelph.ca)

Time

Tuesday & Thursday 10:00 am – 11:20 am

Location

MacN 318

Course Materials and Content

Required textbook

• A. Gezerlis, Numerical Methods in Physics with Python (Cambridge University Press, 2020) The book is freely available to U of  G students: https://ocul-gue.primo.exlibrisgroup.com/permalink/01OCUL_GUE/mrqn4e/alma9953282523105154
• See also the companion website: https://numphyspy.org 

Lecture Content

This is a special-topics course on what is known as computational science or scientific computing. We will focus on the interplay between science problem, mathematical formulation, and computational implementation. Previous exposure to Python programming is required. My current plan is to discuss selected aspects of:

• Floating-point numbers
• Automatic differentiation
• Eigenproblems and the SVD
• Multidimensional minimization
• The fast Fourier transform
• Bayesian and nonlinear regression
• Gauss-Legendre quadrature
• The Metropolis algorithm
• ODEs and PDEs

Course Policies

Expected Background

I expect that all students will have some familiarity with basic numerical methods (e.g., Gaussian elimination, Simpson’s rule, Euler’s method), typically provided in an undergraduate course on computational physics. Programming-related examples and assignments will be in Python, but I will not cover basic programming in the lectures. This is a graduate course, so the assignments will be correspondingly challenging.

Grading

The course evaluation will consist of homework assignments (some of which may not be graded) worth 60% of the final mark and a scheduled final exam (worth 40%).

Office Hours

By appointment.