Instructor: Josh Deutsch 219 ISB

Office Hours: Monday 2:00-4:00 PM

List of topics (tentative)

Computer simulation techniques applied to condensed matter physics such as pattern formation in magnetism and demixing. Monte Carlo techniques, molecular dynamics, and Langevin dynamics.

Other possible topics are: introduction to quantum computing, quantum dissipation, quantum chaos, random matrices. Phonons, semiconductors, and PN junctions.

Books and lecture notes:

Condensed Matter Physics by Michael Marder

Entropy, Order Parameters, and Complexity by J. Sethna

"Statistical Physics'' Richard Feynman.

John Preskill's lecture notes on quantum computing and other resources can be found on his course website

Other links to useful notes on quantum computing can be found on Peter Shor's website

Example scipy scripts:

Here is a tarball of scripts to simulate the Ising and xy models using Monte Carlo. They use weave and blitz to speed up the code by having C code embedded in the scripts. Here is a conseverd.py using conservative (Kawasaki) dynamics.

Here is a tarball of scripts to simulate langevin dynamics for two simple cases, a free particle with mass, free.py, and an overdamped particle connected to a harmonic spring, harmonic.py. You can turn on 3d graphics inside harmonic.py. These employ Visual python. It also computes a histogram of the positions of the particle.

Here is a scipy script for simulating the spreading of a wave packet for the tight binding model in 2d. This also illustrates the use of a class for computing expectation values.

Here is a tarball of a parallel tempering class "ParTemp.py" and its application to a 2d Ising spinglass. For a comparison, here's a tarball of parallel tempering code in C and C++. Here is a tarball illustrating the solution of ODE's using scipy. There is a simple harmonic oscillator script, and also a chain of springs that can be made anharmonic, with analysis of energy sharing between modes. This latter problem was one of the first problems studied on a computer by Fermi, Pasta and Ulam.

Here is a tarball of the "Hopfield Model" a model based on the Ising spinglass, for associative memory.

To run these you'll need python and scipy installed. After typing "ipython -pylab" if all goes well, you should get a prompt and then type "run name_of_script.py"

Things to try using scipy:

Pattern formation links::