The Spring 2009 MATLAB Programming Contest is under way.
As we said, this contest is structured very differently from our usual contests. The complete rules are available here, but in a nutshell, this is an open-ended contest where you are the judge. We want you to look at one very interesting data set and see what kind of insights you can mine from it. The data set in question is one of the most interesting that I know of: the data from one of our previous contests, the Peg Solitaire contest that we ran in May of 2007. I’ve been looking at data from this and other contests for years now, and I’m constantly finding new stories buried in the data. And since many of you actually participated in that contest, you’ll be particularly well positioned to find the good stories hiding in there.
As a side note, you will see a few MathWorkers like me contributing entries during this contest, but we’re not allowed to win any prizes.
What’s going to happen on the main File Exchange page, where every tag is listed individually in the ‘tags’ column, when a file gets hundreds/thousands of vote tags during the contest?
In each files’ page, the repeated tags are listed as vis2009(N). Perhaps this change needs to be made to the File Exchange page before things get out of hand? It would also make it easier to see how many votes a given submission has when browsing or searching by tag.
Although I agree it’s better than Marco’s usage of matlab, this contest is going to pollute the whole File Exchange! And it almost looks like this is intended policy by TMW …
I took a look at the contest data supplied, and noticed that there are no darkness submissions in there (although there is a “darkness” field for the submission data). Is this correct?
You are correct. That’s a mistake. The “darkness” field should be 1 up through d(373) and 0 thereafter. I’ll re-upload the data, but here is some code that will fix your data.
for i = 1:length(d)
if i < = 373
d(i).darkness = 1;
d(i).twilight = 0;
d(i).daylight = 0;
elseif i <= 980
d(i).darkness = 0;
d(i).twilight = 1;
d(i).daylight = 0;
else
d(i).darkness = 0;
d(i).twilight = 0;
d(i).daylight = 1;
end
end
Eric, good point about redundant tags. We’ve implemented a change for this which will go out in the next release of the File Exchange application.
Leave a Reply
About
The MATLAB Programming Contest is a semi-annual competition where contestants submit MATLAB code to try to solve a challenge. For more information, see the overview.
What’s going to happen on the main File Exchange page, where every tag is listed individually in the ‘tags’ column, when a file gets hundreds/thousands of vote tags during the contest?
In each files’ page, the repeated tags are listed as vis2009(N). Perhaps this change needs to be made to the File Exchange page before things get out of hand? It would also make it easier to see how many votes a given submission has when browsing or searching by tag.
Although I agree it’s better than Marco’s usage of matlab, this contest is going to pollute the whole File Exchange! And it almost looks like this is intended policy by TMW …
I took a look at the contest data supplied, and noticed that there are no darkness submissions in there (although there is a “darkness” field for the submission data). Is this correct?
You are correct. That’s a mistake. The “darkness” field should be 1 up through d(373) and 0 thereafter. I’ll re-upload the data, but here is some code that will fix your data.
for i = 1:length(d) if i < = 373 d(i).darkness = 1; d(i).twilight = 0; d(i).daylight = 0; elseif i <= 980 d(i).darkness = 0; d(i).twilight = 1; d(i).daylight = 0; else d(i).darkness = 0; d(i).twilight = 0; d(i).daylight = 1; end endEric, good point about redundant tags. We’ve implemented a change for this which will go out in the next release of the File Exchange application.