Programming C, C++, Java, PHP, Ruby, Turing, VB
Computer Science Canada 
Programming C, C++, Java, PHP, Ruby, Turing, VB  

Username:   Password: 
 RegisterRegister   
 #5?? - DWITE round #2 30th Dec 2011
Index -> CompSci.ca, Contests -> DWITE
Goto page Previous  1, 2
View previous topic Printable versionDownload TopicSubscribe to this topicPrivate MessagesRefresh page View next topic
Author Message
hoyifung04




PostPosted: Fri Dec 02, 2011 9:02 am   Post subject: Re: #5?? - DWITE round #2 30th Dec 2011

Anyone else notice that line 18 of the "fixed" test cases claims that there will be 100,000 lines, yet line 100,019 is still giving us input? It is not until line 100,469 that there is another number, meaning that there is another 450 extra lines after the "end" of the third data set and before "start" of the fourth data set.
Sponsor
Sponsor
Sponsor
sponsor
hoyifung04




PostPosted: Fri Dec 02, 2011 9:29 am   Post subject: Re: #5?? - DWITE round #2 30th Dec 2011

lufthansa747 @ Thu 01 Dec, 2011 21:58 wrote:
i found a problem this line in the test cases is missing a space

p eelhieluaelldyplb[...]


This isn't the only case like that. There are a total of 368 lines (including that one) that only have one word (instead of two). All of them are in the large test case with long names (number 4).
A.J




PostPosted: Sat Dec 03, 2011 5:54 pm   Post subject: RE:#5?? - DWITE round #2 30th Dec 2011

@crossley7- Yeah, I'll do anything to make sure that NEVER happens again Very Happy
crossley7




PostPosted: Sat Dec 03, 2011 8:35 pm   Post subject: RE:#5?? - DWITE round #2 30th Dec 2011

I don't mind worst case scenario in terms of mathematical obscure cases, but the time out type massive cases are a pain. When the entire time is spent doing read/write, you have to have a negative run time solution Very Happy
PatrickWhite




PostPosted: Sat Dec 03, 2011 9:30 pm   Post subject: Re: #5?? - DWITE round #2 30th Dec 2011

What crossley said.
An example of a good limit would be the accidentally easy question from last time, amended with Cyril's limit. That would kill off bad algorithms, not good algorithms that take time to read files.
A.J




PostPosted: Sun Dec 04, 2011 12:06 am   Post subject: RE:#5?? - DWITE round #2 30th Dec 2011

That's just the problem. There are a lot of problems that kills off bad algorithms just because of the time constraints. In fact, a lot of path finding questions (well, some of the more challenging ones at least) can be solved using a more naive implementation if the input constraints weren't too strict.

I realise that the big testcases did cause problems for this round, so we'll keep these type of questions to a minimum. However, if we want to incorporate non-standard questions, we might have to increase the input constraints once in a while.
SamScott




PostPosted: Sun Dec 04, 2011 2:44 pm   Post subject: Re: #5?? - DWITE round #2 30th Dec 2011

I have downloaded the test file, removed the underscores, fixed the line with no space in it as above, but I have found many more just like it. I started fixing them in the way you suggested - making sure they form a big chain - but there are just too many of them to fix by hand. Other than that, my answers are agreeing with yours.
nayuki




PostPosted: Sun Dec 25, 2011 10:57 pm   Post subject: Re: #5?? - DWITE round #2 30th Dec 2011

Here's my attempt at fixing the input data. I hope this will be the last.

File: https://github.com/nayuki/DWITE-programming-contest-solutions/blob/master/testdata/201111-DATA5.txt
Sponsor
Sponsor
Sponsor
sponsor
Display posts from previous:   
   Index -> CompSci.ca, Contests -> DWITE
View previous topic Tell A FriendPrintable versionDownload TopicSubscribe to this topicPrivate MessagesRefresh page View next topic

Page 2 of 2  [ 23 Posts ]
Goto page Previous  1, 2
Jump to:   


Style:  
Search: