Numerical Solutions of Linear Equations

Given the linear system Ax=b, we often represent the solution as x=A^-1b. But this does not imply you should actually compute an inverse when you need to solve a system. John Cook offers a short explanation of why you should avoid inversion. For example, if you import NumPy's linear algebra module as la, then you can use la.solve(A,b) instead of la.inv(A)*b.

Introductory Announcements

There will be an introduction to Python programming in the SSRL on Friday, August 27, from 5:30-7:00pm. Please be punctual.
Course materials, including the syllabus and course notes, are available via Blackboard. Be sure to sign up for the class mailing list by following the directions on the syllabus.
All software used in this course should be available in the SSRL. If you want to work at home, please install the free and open source Python 2.6.5 and Numpy 1.4.1. (You can also find these bundled together in the Enthought Python Distribution.) Additionally, you will need the commercial Scientific Notebook, which you will use to write up your homework assignments. (SN is reasonably priced, but MacKichan additionally offers cheap short-term licenses.)

Monkeynomics

Laurie Santos's 2010 TED talk provides some interesting insight into the biological basis of human decision making strategies. Dr. Santos is a bit quick to call behavior “irrational” when it deviates from classical predictions, but that does not diminish the interest of her results.