The Nineteenth Byte

The Nineteenth Byte: General discussion for codegolf.stackexc...
May 22, 2019 11:30
May 20, 2018 08:00
Thanks. Finally some good results this year.
Btw, also thanks for the scraped scores on you webpage.
 
May 15, 2017 06:02
todd_and_steven gave a wrong answer, and weird_editor had a runtime-error. Still don't know, why any of these happened. I was very confident with both solutions.
May 15, 2017 06:01
Yes, quite odd indeed.
May 15, 2017 06:00
@aditsu Thanks
May 14, 2017 17:35
The DCJ-team said, that it will take several hours to judge the large submissions.
May 14, 2017 17:31
@aditsu Might be. But it will be tough. Quite a lot of people have 67 points currently.
May 14, 2017 17:30
Also did some binary search stuff, but a little bit different. Each node i exactly computes the result for [ilength/100, (i+1)*length/100-1]. For this I did a binary search (that itself uses a binary search) to compute the value A[ilength/100] (A is the merged array).
May 14, 2017 17:18
Not sure.
But I just realized that I made an error, that will probably my program to exceed the message memory limit. So it will probably not work.
May 14, 2017 17:13
And for large I tried something similar. Split on 98 nodes, report the bad interval to node 98, this one computes the bad index and sends it to 99, which does (with the help of the first 98 nodes) the sum computation.
May 14, 2017 17:12
Yes, this idea would work.
For the small case, I simply called GetValue(i) ten times for each i. Once these 10 values aren't equal, I have found the bad index and send it to the second node. And this one sums all values (with the bad index at the end).
May 14, 2017 17:07
Btw, Congrats on winning a shirt yesterday.
May 14, 2017 17:06
On the terminal it still says that I don't catch all messages.
May 14, 2017 17:05
@aditsu I don't belive that my E-large will work.
Apr 30, 2017 11:36
@Sp3000 Actually not. Failing A-large only gives you place 1016.
Apr 30, 2017 11:31
Congrats
Apr 30, 2017 11:15
Assuming that a few other large ones fail.
Apr 30, 2017 11:15
You might only need 1 large to qualify.
Apr 30, 2017 11:05
and aditsu also just got 72 points.
Apr 30, 2017 11:05
cheers
Apr 22, 2017 21:10
Nice
Apr 22, 2017 20:55
*bug
Apr 22, 2017 20:55
Or is this a but?
Apr 22, 2017 20:55
@aditsu I've notices, that some guys didn't participant in the qualification round. For instance go-hero.net/jam/17/name/alexmit. Only 1A and 1B. Don't you have do "qualify" in the qualification round?
Apr 22, 2017 20:47
Yeah, small country.
Apr 22, 2017 20:44
Thanks
Apr 22, 2017 20:44
I'm responsible for every language except C++, and I'm doing C++ in the next rounds.
Apr 22, 2017 20:44
Languages used by the top 20% of Austria.
Apr 22, 2017 20:43
Apr 22, 2017 20:43
Stats updated.
Apr 22, 2017 18:48
That could have save me a lot of time in the qualification round.
Apr 22, 2017 18:48
Oh, the cjam-interpreter has a code jam modus.
Apr 22, 2017 18:42
Used ruby for the B-small. And I had the biggest problems with ruby. Couldn't even figure out how to use the sort function. So I sorted it by comparing the elements and swapping them...
Apr 22, 2017 18:40
3 more
Apr 22, 2017 18:38
Luckily I found the code on the codegolf-tips page.
Apr 22, 2017 18:38
@aditsu Yeah, figured that out after ~40 minutes.
Apr 22, 2017 18:36
Oh, that's small. My solution is a little bit larger:
{2.-1??./*}:float;
[~]
(;0\
{
(\0\(
{\
(\(@ 5$\-\float/.4$>\4$if@@+@;(
1-.}do
;@@/
@)@+"Case #"@@(@@": "@@"\n"\4$\
.}do
;;
Apr 22, 2017 18:34
It was probably was not a good idea to write A-large in golfscript. Took me over an hour.
Apr 16, 2017 17:53
When this is possible it is not obvious which interval you should pick. But of course this is not possible. If the quantity A is smaller than quantity B, then interval(A).start <= interval(B).start and interval(A).end <= interval(B).end. So just sorting and greadily picking the smallest is enough.
Apr 16, 2017 17:53
@aditsu: Oh, that's actually more easier than my solution. Somehow I thought that there can be two kit-intervals of the same ingredient, that are strangely overlapping. E.g. the first package contains enough tomatoes for [3-4] servings and the second package contains enough tomatoes for [2-5] packages.
Apr 16, 2017 08:46
@KritixiLithos No, just a private person.
Apr 16, 2017 08:44
This is how I solved it. The official solution is pretty similar. For each ingredient Qij in each package you can compute the number of kits, that you can create using it. For instance if the optimal recipe requires 100g of tomato and you have 1100g, than you can create 10, 11 or 12 kits. Notice this will always be an interval, in this case [10, 12]. You can compute the interval boundaries using a little bit of math.
After converting each Qij to an interval, you want to find intervals, that overlap. E.g. you can combine the onion interval [9-10] with the tomato interval [10-12], because 10
Apr 15, 2017 18:54
@KritixiLithos You can use paid languages in the qualification round. Only in the "real" rounds they are banned.
Apr 15, 2017 17:12
Wow, currently 3rd in multiple languages.
Apr 15, 2017 17:06
Finally. Took him a week to change it.
Apr 15, 2017 09:38
@aditsu Yes. "The Wolfram Language is the programming language used in Mathematica." (Wikipedia)
Apr 15, 2017 03:45
I'll use C and C++ in the next rounds. And maybe Golfscript, if there is a opportunity (input that doesn't require obnoxious parsing, ...).
Apr 15, 2017 03:40
But I probably should have saved Python and Java for a serious round.
Apr 15, 2017 03:40
JavaScript, D, Python and Java
Apr 15, 2017 03:39
Used 4 more. ;-)