UP | HOME

Assignment 4 Speed Contest Leaderboard

All tests were run on a Mac Pro 12-core desktop computer running OS X 10.10.1 and gcc version 4.9.2, MATLAB R2014b, Python 2.7.8 (Anaconda 2.1.0) and R version 3.1.1

Once everyone hands in their code I will make the code available on this webpage.

You may resubmit your code any time up to the deadline (e.g. if you think of a new way of speeding it up after handing it in).

You're also free to submit code in multiple languages if you want to compete in more than one competition.

name language speed (sec) parallel notes
Gribble C 0.498 Y 1
Gribble C 4.150 N 1
Gribble C 7.920 N  
Dekraker MATLAB 8.753 Y 1,2
Gu MATLAB 42.449 Y 1
Dekraker MATLAB 43.005 Y 1
Nguyen & Vo MATLAB 52.446 Y 1
Ritz MATLAB 66.176 Y 1
Abeyasinghe MATLAB 82.674 Y 1
Sternin MATLAB 106.869 Y 1
Lyons & Sergeeva MATLAB 110.293 Y 1
Viczko MATLAB 122.403 Y  
Gribble MATLAB 124.953 Y  
Blumenthal MATLAB 1137.043 N 1
Domingo & Tran MATLAB 1143.310 N  
Lyons & Sergeeva MATLAB 1150.420 N  
Gribble MATLAB 1151.175 N  
Gribble Python 2526.500 N 1
Coros & Alazary Python 6106.359 N 1
Gribble R 95568.750 Y 1,3
Nichols R 3330647.000 N 4

Notes

  1. Made platemethod function code more efficient
  2. Used MATLAB compiler to generate compiled versions of the platemethod function and a function for doing the bootstrap
  3. Timed on 24 neurons and 100 bootstrap iterations and extrapolated the time required for 100 neurons using 10,000 iterations
  4. Timed on 100 neurons and 10 bootstrap iterations and extrapolated the time required for 100 neurons using 10,000 iterations

last updated Dec 20 2014


Paul Gribble | fall 2014
This work is licensed under a Creative Commons Attribution 4.0 International License
Creative Commons License