Computational Physics – Overview

Contents

Schrödinger equation / object oriented programming; atoms in field / coordinates and symmetries / managing complex codes; Ising model / random numbers and quantum Monte Carlo; classical mechanics and chaos / time-integration / program modules; linear solvers / scattering problems / LAPACK and FFTW; methods in many-particle systems / parallel computing
Codes will be written in Python and (optionally) C++.

Goal

At the end of the course you will be able solve complex physics problems on the computer. You will understand key numerical issues, you will have learned modern object-oriented and parallel programming.

Prerequisits

Quantum and classical mechanics, numerical methods, basic programming experience (MATLAB, C, C++, Fortran, or Python). http://www.mathematik.uni-muenchen.de/~kerscher/vorlesungen/numeriksose10/doc/script.pdf

Registration

Please, register for lecture and C++ course through the LSF online "Vorlesungsverzeichnis". Registration facilitates communication and administration. Further it will allow us to give you access to the SVN server to be used for lecture and C++ course.

C++ Lab course

In addition to the regular Übungen an intense computer lab course (2 hours/week) will be offered. This course will closely follow the lecture, but implement the presented problems in professional level C++.

Useful literature

• Puls, Stintzing, Kerscher: Numerik für Physiker (lecture notes)
• Thijssen: Computational Physics
• Golub, vanLoan: Matrix Computations
• Press,Teukolsky,Vetterling,Flannery: Numerical Recipies
• Überhuber: Computer-Numerik
• Lutz: Learning Python
• Prata: C++ Primer Plus

Crediting of the course

The course will be credited when the written exam at the end of the course is passed successfully. Details will be announced in due time.

Contact

Professor: A. Scrinzi
Room: A206
Phone: 089 / 2180-4608
armin.scrinzi@lmu.de