Project Euler Forum

Can someone confirm these results for Problem 233 (View Problem)?

Some random values of N such that f(N) = 420: 1328125, 84246500, 248431625.
My code compute approx 10^7 such numbers <= 10^11.

Thanks in advance.

SmuK wrote:Some random values of N such that f(N) = 420: 1328125, 84246500, 248431625.

f(1328125) = 180

f(84246500) = f(248431625) = 420

"Congratulations, the answer you gave to problem 233 is correct."

Thanks a lot. It was a stupid bug

My code returns 30875234922 for n<=38000000, is it correct?

Yes.

for n<=10^10 i have 2709499279563106.
is it correct?

quintana wrote:for n<=10^10 i have 2709499279563106.
is it correct?

Unfortunately, it's not.

You aren't the first one to get this particular result, though.

it's interesting because i have correct answer for 38000000. thanks!

Yeah, you're on the right track, but are missing something that a lot of other people missed initially too.

quintana wrote:it's interesting because i have correct answer for 38000000. thanks!

It seems you got it now

Congratulations!

thanks, i got it allright! i got the idea long time ago, but like p180 this one had a catch beyond idea. again, forum was very helpfull.

Just a quick question for clarification:

"sum of the integers": Does it mean sum of the coordinates, that are integers up to 10^11? And if so: X or Y

What is the sum of all positive integers N

The actual value of the other integer coordinates on the circles are immaterial. Look at the definition given for f(N).

quilan wrote:Yeah, you're on the right track, but are missing something that a lot of other people missed initially too.

I have an elegant algorithm that gives the same wrong results for 1e10, 1e11 and the correct one for 38e6. I don't see any way it could be double-counting solutions.

Now I'm running a brute-force check counting down from 1e10 to see where it's making the mistake. It's taking forever. AAARGH! Would it help if I pull my hair out?

Seriously, this isn't something dumb like an overflow bug, is it?

update: Got it! It wasn't an overflow bug, I just had to look a little deeper into the characteristics of f().

Same as everyone. As for now, i can't see any difference for even and odd N. And i can't understand why answer for 38000000 is correct and answer for 10^11 is not. I simply don't see any special cases between this numbers.

My code finds 5422051 solutions for 10^11. Is it too big or too small?

Update: Wait a bit, I've got an idea.

Update 2: And hero have won once again without any help!
There are 5422629 solutions in total and there is no difference between even and odd N. It's something else.

The subject of N <= 10^10 came up, but without a definitive response. Can anyone tell me if the correct answer for 10^10 is...

2709680426923506

It's very close to what quintana had, but not quite the same.
I'm getting the right answer for N <= 38 million (though I'm not sure what was special about the number 38 million), but evidently not the right answer for N <= 10^11, and it's driving me a little nuts.

Thanks!

Sorry, it's too small.

Sorry, small bug.

for N <= 10^10: 2710041165705856 ?

EDIT:
Nevermind! Introduced another small bug while fixing the first. Now they're both fixed!

Thank you for all the very helpful comments on this question. It helped me a lot when I saw a lot of other people missed one little thing. After I found that I felt I should give my diploma back.
So far one of the best problems I solved, even if it took me three days to solve it, but it worth.

Thanks to all posters above. I have to say that the open forum was great for this one - thanks for giving a sample answer for the problem on toy scale. When I realised my answer for 38000000 was wrong, i went hunting...

Helped me find a stupid bug using a less than instead of less than equals!!!! I'd had the code right since first writing it but for this bug!

Probably the hardest one I solved so far. Got the idea quite quickly, but debugging was a nightmare.