Hello guys!
We gladly invite everyone to participate in International Online Programming Contest (IOPC), 2016, organized by Techkriti, IIT Kanpur. The problem setters and testers are FacelessMen (alecsyde, sahilgrover, abhilak) and AKS++(TerryMcGinnis, webmail, freaken). This is our first time setting a contest and we have tried our best to make an interesting problemset. The contest will be 24 hours in duration and you can see when the contest starts in your timezone here. It is a team contest. It starts tomorrow, 12th March at 6 PM(IST).
Total amount of prizes at stake here are 120,000 INR (1750 USD). The distribution will be 50% to the 1st position, 30% to the second position and 20% to the 3rd position. The registration link for the contest is hosted here. The theme of the contest will be Batman V Superman.
Edit : The new prize distribution scheme is as follows: Top 3 global teams will have 90,000 INR distributed as 50% to the 1st position, 30% to the second position and 20% to the 3rd position. Top 3 Indian teams will have 30,000 INR distributed as 50% to the 1st position, 30% to the second position and 20% to the 3rd position.
Edit : Delayed by 1 hour, to 12th March at 7 PM(IST).
Auto comment: topic has been updated by TerryMcGinnis (previous revision, new revision, compare).
Prizes for top 20 or so would have made it more exciting, especially considering your budget.
We're discussing it's practicality right now. We'll let you know if something changes. Thanks for the suggestion!
Updated!!
Auto comment: topic has been updated by TerryMcGinnis (previous revision, new revision, compare).
We participated in last year's IOPC and have not received the prize money.
I talked to the Techkriti team about this. They have received a lot of complaints regarding this and even we knew about the incident, our condition to the Techkriti team for making the contest was that the prizes would be delivered this year. I regret that for the previous year, your prize money was not given to you and we cannot do justice to that. :(
I have tried to make sure that you will like the problem set. Happy solving.. :)
what happens to the prize money in such cases?
They could just express apology every year, and never need to deliver the prize.
It happened to me too. In codechef too! What a coincidence. And the problem setter was terribe. He didn't reply my e-mail for all the mistakes, everywhere! I think the problem with Codechef is in the hosted contest. Seems like everyonecan make a hosted contest there, but literally, everyone with no experience in programming! We should be proud and happy for having this nice Codeforces :D
Auto comment: topic has been updated by TerryMcGinnis (previous revision, new revision, compare).
Auto comment: topic has been updated by TerryMcGinnis (previous revision, new revision, compare).
Can't read problem statements, it says: This is a restricted contest. It will only be accessible to users who are provided the permission ...
Is it normal?
When I try to enter the contest it says.. "You will have to login to check the contents on IOPC 2016 contest page" although I am logged in with a team registered for the contest. Anybody else facing the same issue?
It is okay now.
Working fine now. But the timer seems to be incorrect or has the contest been reduced to 23 hours?
It will be extended by some hours. Please wait for some time before we decide the extension time, and update the time there.
Codechef people had some problem from their side. It has been fixed now.
What does it mean?
How to solve "Superman and the 5th-Dimension" ?
My teammate solved this problem, but it seems is like this:
Let's solve the problem for all matrices of width W
If the submatrix starts at column c, then we can calculate the gcd of all the rows of length W that starts at column c
Now the problem is given an array calculate the segment that maximizes length·gcd, that can be done in n·logn because gcd decreases logarithmically
Can you explain how "length·gcd, that can be done in n·log n"? How is "gcd decreases logarithmically used here ?
if the numbers are unique, then the gcd decreases at least in a half, so you only need to start from one position and go to the right until the gcd becomes 1;
I left to you the case when the numbers are not unique ;)
A well prepared, and a great quality contest. ( Nothing unusual about IOPC on that front! ) When will the editorials be published? Also, what was the approach for Collateral Damage?
Thanks a lot! We might not be able to publish fully explained solutions but we will try to post short solutions of all the problems and provide setter/tester code. As far the problem Collateral Damage, basic approach is the minimum cover of a set of circles with same radii is a convex polygon like figure with curves along the vertices and straight lines tangent between consecutive circles. So basically u need to find the convex polygon formed by the centers. The tangents joining consecutive circles and the arcs form the cover. You can compute the arc lengths using simple geometry and add them up but that might cause some precision issues (though we did allow for that) but a nicer observation is each arc length is r*external angle at a vertex and since sum of external angles is 2 * pi, the answer is total length of convex polygon of centers + 2 * pi * r.
How to solve problem BATGODEL and SUPERTURING ?
It's absolutely theoretic. See here. The problem requires only parity of permanent, so it equals to parity of determinant.
How to reduce this problem to the above concept ?
By taking several liberties with the problem statement, apparently. For instance, a graph with two self-loops at 1 and 2 and an edge from 1 to 2 satisfies the property that every vertex belongs to exactly one cycle, yet you shouldn't consider it for some reason...
It seems like you are frustrated with the problem statement. Please answer to my question if you can.
In the problem, we can see that the edge with even weight is useless, so just ignore them.
For each vertex u in the directed graph, split it into two vertices uin and uout, for each edge u → v with odd weight in the directed graph, add an edge from uout to vin. You can see that, the new graph formed by uin, uout is a bipartite. And any perfect matching in the bipartite corresponding to a Turing-Subgraphs. So we just need to counting the parity of the number of perfect matching. This can be done by find the parity of permanent.