|
|
 |
|
|
|
|
|
|
|
8.981 :: Selected Topics in Astrophysics |
|
|
About this class
This class is a seminar on computational techniques of use in physics
and astrophysics research. We will generally follow the excellent
resource Numerical Recipes by Press, Teukolsky, Vetterling, and
Flannery.
My philosophy on why a course like this is important and interesting
closely follows that offered by Professor James Sethna, who teaches a
similar
course at Cornell; I'm going to shamelessly crib his blurb right
here:
There are three main reasons that serious computational physicists and
engineers should know this material, even though computational
environments like Octave, Python, Matlab©, and Mathematica©,
provide "black-box" routines that will reliably and efficiently
perform many of these tasks:
1. The black boxes often fail just where the physics is most
interesting. Knowing how they work is crucial for finding
replacements.
2. For computationally intensive tasks, one can often make use of (or
design new) specialized routines that outperform the general-purpose
routines.
3. Amazingly often, researchers will use their knowledge of algorithms
to apply the basic ideas in a completely new context.
To this, I would add 2 additional items:
4. "The purpose of computing is insight, not numbers." (Richard
Hamming.) Knowing how to crack open the black box and rebuild it
yourself teaches you an extraordinary amount beyond what you get from
cranking out numbers.
5. Building your own routines can be amazingly fun in a way that
really speaks to one's inner (or outer) geek. You wouldn't be taking
a high-level physics course if this didn't appeal to you in some way.
(It should be admitted that this kind of code hacking can also be
amazingly frustrating.)
Lectures
Staff
Lecturer: Prof. Scott Hughes. Office: 37-626C; Telephone 8-8523; sahughes@mit.edu.
Office hours: By appointment. Standing office hours will be
offered if there is sufficient interest.
Homework and grades
We will have weekly problem sets featuring short computational
exercises for the first N weeks of the term; we will move to
student projects and student presentations for the final 14 - N
weeks of the term. N will be determined based on the course's
enrollment.
Note that Numerical recipes C++ routines relevant to many of
the topics we will cover. You are encouraged to use those routines
if that is the mode in which you tend to operate. If you
prefer to use a tool with which you are more comfortable
(Mathematica©, Octave, Matlab©,...) or use scientific
computing libraries (NAG, GNU, IDL, ...) that's fine too.
This course will be graded Pass/Fail.
|
|
|
|
|
