[Community Puzzle] The leaking bathtub

Puzzles
#1

Coding Games and Programming Challenges to Code Better

Send your feedback or ask for help here!

Created by @Tatur2000,validated by @Razovsky,@Diabetos and @sammck.
If you have any issues, feel free to ping them.

1 Like
#3

Passed all tests, can’t pass validator 2 and 5. I use cm per seconds, so when water’s level get to (h-level) I get time needed to fill bath, then translate seconds to hours and minutes.
Types of variables are double (C++).

#4

Have done by another method: counted time in depends of baths volume devided on filliing volume till leaks point.
Passed all tests, but didn’t pass validator 3

#5

Hello, great problem so far! I had the same first approach as @Oliverri , very inefficient, I got performance issues on the last test. This made me rethink the problem and I found a much more elegant solution. I think its pretty much the same as Oliverri’s second method actually.
I pass all the tests but not the first validator, and I don’t get why. Do you have some hints? Thanks! :slightly_smiling_face:

EDIT: oh, I think I got something. Now I’m computing minutes and deducing seconds and hours from it, for test 2 I get 17.40926 minutes, so 17 minutes and 0.40926 * 60 = 24.5556 seconds. I round to 25 but the expected answer is 24. Why is that? Am I losing precision somewhere?

#6

Description says that result should be rounded down so 24.5556 should be displayed as 24 seconds

#7

Hello Yatech, thanks for noticing. Okay, so rounding down is the same as truncating the floating part, which is what I’ve done before coming to this post: when truncating, I pass all the tests but not the first validator and I don’t quite get why…
It looks like I’m missing a whole second, because if I submit a code that adds 1 second to my existing code, the first validator passes.

#8

Most likely it’s the problem with precision. Try using double instead of float :slight_smile:

For validator 1 the total minutes have decimal part .333333333333336.
If the type you are using is not precise enough this will turn into 19 seconds instead of 20.

#9

Thanks for the insight, that was a precision issue indeed :slight_smile:
However I didn’t have to use double instead of float, because such a piece of code:

float m =  0.333333333333336f;
int s = (int)((m - (int)m) * 60);
Console.WriteLine(s);

Returns the expected result (20).
So instead, I used a comparison like this:

int seconds = (int)((minutes - (int)minutes) * 60);
float secondsF = (minutes - (int)minutes) * 60;
if (Math.Abs(secondsF - (float)(seconds + 1)) < 0.001f)
{
    seconds++;
}

Seems to do the trick. Let me know if you have a better solution because it still feels wonky…
Anyway, thanks again!

#10

My C# code is published if you want ot take a look.
I just used double and formatted the answer using TimeSpan.