I saw someone posted a code snippet about Pascal's triangle, so I tried.

def my_pascals_triangle(n):
  d = [0, 1, 0]
  yield filter(None, d)
  for i in range(3, n + 2):
    # i == len(d)
    d = [d[j - 1] + d[j % i] for j in range(i + 1)]
    yield filter(None, d)

The performance is better in Python 2.5/2.6. The code in that post took 0.62 seconds for n=1000, mine took 0.43 including converting to list. I made another version from his code, so it could also using generator way:

def pascals_triangle_1(n):
    x = [1]
    yield x
    for i in range(n-1):
        x = [sum(i) for i in zip([0]+x,x+[0])]
        yield x

The performance is same, around 0.62 seconds including converting to list.

However, if I use Python 3.1, mine got pretty bad performance, 1+ seconds. His even gets better, 0.45 seconds.

I bet some Python list comprehensionist might be able to write it in one-liner without using additional defined function and with standard Python libraries if necessary. If you are the one, show off, please!