### jakubd's blog

By jakubd, history, 3 months ago,

people who competed in 2011 are laughing so hard now on >1900 rated people crying about FST on Div. 2 C lmao

• -30

By jakubd, history, 4 months ago,

How come there is no European Boys Olympiad in Informatics? It's not like girls are limited by nature to perform as good as boys, as in sports? :thinking_emoji:

• -21

By jakubd, history, 4 months ago,

Why can't we just register for contest whenever? I know it's my fault for not clicking register ahead, not the first time lol, but bruh, why do we have to wait 10 minutes into contest when we forgot to click register button more than 5 minutes before contest?

• +46

By jakubd, history, 6 months ago,

Why do some grader's count the last submit and not the best submit (in tasks with points)?

Is there any advantage to this?

• -19

By jakubd, history, 6 months ago,

• +17

By jakubd, history, 6 months ago,

Hello everyone.

I was solving some past COCI problems, to be exact problem 3. Maraton from the 6th contest of 2006/2007 season and could not get a full score, even the 2nd sample could not pass, where you should just output the word "ongoing". After a couple of attempts I even submitted the official solution code and even that only got 27/30 points. (1 test case failed and also the mentioned sample).

Does anybody know, who I could contact about this? Because there is no contact page on the evaluator website. Any help would be appreciated. Thanks

• +26

By jakubd, history, 7 months ago,

The problem is you have n line segments, where n is upto 10^6 and you should count least amount of rays starting from origin (0, 0) necessarry, so that each line intersects with at least one ray.

The solution to the problem is radial sweep. First sort the endpoints of the line segments by angle and then do a simple sweep, in which if we encounter the end of some line, we do eitherbof two things:

• create a ray which goes at a angle of the encountered endpoint from origin and we also increase the answer by 1.

• we have already created ray with angle between the angle of start of the current line segment and the end of the current line segment, in such case we do nothing since there is already a ray going through that line segment

And in the last sentence of the second scenario is where my program fails. Sometimes you should cast a ray, for example if you have following two lines and you are on point D in your sweep and have already gone through points CBA in that order:

In this moment my mentioned program decides to not do anything, since the angle st which we have casted a ray through point A is between angles of point D and C, if we sort using the function atan2l.

Unfortunately, I cannot figure out how to solve this edge, case. I have, by hand, found out that there are numerous such cases, but I can't make my program solve them and I can't even describe these cases or group them by some characteristic which would help me solve them.

Any help with that would be aprreciated. Thank you very much.

• +13

By jakubd, history, 7 months ago,

It should say Became Expert, but it does not for some reason (maybe a magic related bug?). Has this ever happened to you?

• 0

By jakubd, history, 7 months ago,

Correct options:

• 4

• 2

Wrong options:

• +13

By jakubd, history, 13 months ago,

Hello everyone.

Just got a message that my submission from problem A of the last Div.3 Round is very similar to these two other submissions: 121915999 and 121935052.

I have no reason to cheat and I definitely don't have to do it on problem A in a Div.3 contest as you can see on my results from all the previous contests I participated in, so my question is: What can I do to get my results counted?

Thanks for the help.