Read jiangly 's code. (I don't dare to tag him.)

one suggestion is: when you find a problem that you try but can't implement, search other people code for the same problem and try to understand how they solved it, so when you find other problem like that you will at least have some idea to how implement it.

When you are saying "my idea", what exactly do you mean? For example, if your idea was something very vague like "this appears to be a DP problem" and later you read in the editorial that it was indeed a DP problem, then you can't say that you had the right idea. Was your idea actually detailed enough to start implementing it or are you basically deceiving yourself?

Or do you have a correct idea (later confirmed by the editorial), attempt to implement it, get wrong answers and then fail to debug your code?