Nov 27
I've put more here that we can do today.
We'll pick and choose and/or continue Thursday.
Options
-------
1. Final project questions?
2. Course evaluations (Tue or Thu)
3. Big term review.
4. Examples:
4a) scientific python
4b) Tron google competition (aichallenge.org , winter 2010)
4c) traveling salesman problem
1
Final project questions? Presentations one week from today; due a week from Friday. Discuss status.
2
Course evaluations.
3
Review: walk through everything we've done this semester ... with your help.
4
Python examples :
scientific python
Discuss numpy, matplotlib, etc; look at some examples.
Fast manipulations of big collections of numbers,
including plots, transforms, programming.
Similar tools: MatLab, IDL, GDL
Numpy has several classes for collections of numbers:
- arrays (element by element operations)
- matricies (matrix
While many people use python for serious numerical work,
it is unfortunately true that getting all the pieces
installed isn't always simple. For one thing, the
current packages are a merge of several others,
including numpy and scipy; for another, many of the
libraries depend on C and Fortran compilers that
don't always install easily. (Apple, for example,
is now using their own C compiler llvm , which
doesn't always behave the same way that gcc -
what everyone else uses - behaves.)
One effort to build an "easy to install all-in-one"
package is this one, which I'll use a demo here.
$ python
>>> from pylab import * # all matplotlib.pyplot & numpy names
# plot of sin(x)/x
>>> dx = 0.01
>>> x = arange(dx, 4*pi, dx) # array range
>>> y = sin(x)/x # element by element
>>> plot(x, y)
# plot of a histogram of randoms
>>> randoms = randn(1e4) # 10,000 gaussian randoms (mean=0, variance=1)
>>> hist(randoms, 100) # graph 'em in 100 bins
# statistics
>>> mean(randoms) # near 0
>>> std(y) # standard deviation (near 1)
# manipulating an array
>>> x.ndim
1
>>> x.shape
(1256,)
>>> x[3:4]
array([ 0.04, 0.05, 0.06, 0.07, 0.08, 0.09, 0.1 ])
# matrix work
# *** note that numpy's, nm[i,j] is python's m[i][j] ; i=row, j=column
>>> m = matrix([[1,2,3], [1,0,1], [0,2,-1]])
>>> m
matrix([[1, 2, 3],
[1, 0, 1],
[0, 2, -1]])
>>> m[0,:]
matrix([[1, 2, 3]])
>>> m[: , 0] # remember that a[:] is all elements
matrix([[1],
[1],
[0]])
>>> m * m # matrix multiplication
matrix([[ 3, 8, 2],
[ 1, 4, 2],
[ 2, -2, 3]])
>>> m[0:2, 0:2]
matrix([[1, 2],
[1, 0]])
>>> inv(m)
matrix([[-0.33333333, 1.33333333, 0.33333333],
[ 0.16666667, -0.16666667, 0.33333333],
[ 0.33333333, -0.33333333, -0.33333333]])
>>>
Tron robots
from a previous year's programming workshop
Browse through a bit of the code:
Run an example from notes on my laptop at
/Users/mahoney/academics/term/2010-01-spring/programming_workshop/tron/
traveling salesman problem
Some code I wrote while looking at a classic CS problem
with a plan student several years ago (attached).
![[paper clip]](/courses/source/wiki_images/paper_clip_tilt.png)
name last modified size