I still wonder how people get this calculated in under 0.002s..
What worked for me was to first cache pairs (c, d) for every x: 1..20'000'000 such that
x = 2^a * 5^b * r;
c = a - b; // easily calculated
d = r % 10;
Then considering:
if x -> (c, d) then lastnonzerodigit(x) = { d, if c = 0 | 5, if ...
Search found 3 matches
- Mon May 05, 2008 2:21 pm
- Forum: Volume 102 (10200-10299)
- Topic: 10212 - The Last Non-zero Digit.
- Replies: 63
- Views: 41353
- Mon May 05, 2008 11:25 am
- Forum: Volume 102 (10200-10299)
- Topic: 10228 - Star not a Tree?
- Replies: 4
- Views: 3248
Re: 10228 - A Star not A Tree
Gradient descent with centroid as a starting point works well too.
Is there a geometrical solution though?
Is there a geometrical solution though?
- Mon May 05, 2008 11:08 am
- Forum: Volume 102 (10200-10299)
- Topic: 10246 - Asterix and Obelix
- Replies: 29
- Views: 14650
Re: 10246 - Asterix and Obelix
A simpler example where single Floyd-Warshall is too greedy: it discards shorter path over city 3 because it's expensive to party there, even though a more expensive city (5) is inevitably encountered later.
5 5 1
10 10 20 10 30
1 2 2
1 3 1
2 4 2
3 4 1
4 5 5
1 5
I'm guessing running the ...
5 5 1
10 10 20 10 30
1 2 2
1 3 1
2 4 2
3 4 1
4 5 5
1 5
I'm guessing running the ...