Jim & Jaggo - Dict lookup is (essentially) constant-time because the hashing function computes which unique bucket a given entry will correspond to. (Hashing functions are usually polynomials involving prime numbers. Can assume that the computation of the hash value is constant-time)
So there is no linear table-walking in dict lookup. Walking the shorter list will thus be faster in general, esp if M<<N. It would be neat to code this example up and measure performance using the standard timeit.Timer object (does anyone have any code or package to generate GIF graphs with logarithmic axes for N, based on timeit.Timer output? Maybe Matplotlib?) If Jaggo could suggest us typical ranges for M,N in his application that would help guide the discussion... cases like M=10, N=10^6 will showcase this. (I am real busy on another task I don't have time to do it right now) PS What is the name of the author of the 'profiler' module? Stephen _________________________________________________________________ Puzzles, trivia teasers, word scrambles and more. Play for your chance to win! http://club.live.com/home.aspx?icid=CLUB_hotmailtextlink _______________________________________________ Tutor maillist - Tutor@python.org http://mail.python.org/mailman/listinfo/tutor
