This is O(n^3) because of the goto statement (effectively you have replaced a loop with goto :) I think. Instead of neightbor in C, you can do a binary search. This is what I could think of but it doesnt meet the problem's requiremen - O(nlgn): 1) O(n^2logn) with O(1) space:) Sort all 3 arrays. For each pair of a[i],b[j], binary search for (0-(a[i]+b[j])) in C to see if its present. 2) O(n^2) with O(n) space Same as above, but instead of binary search of 0-(a[1]+b[j]) on C, you put all elements of C in a hash. There must be some variation of merge sort to do it in nlgn, but Im not able to see it :)
On Sun, Nov 1, 2009 at 12:39 AM, daizi sheng <daizish...@gmail.com> wrote: > with all arrays sorted firstly, if you enumerate ai, bj in ascedning order, > ck will be sure in descending order. > > foreach(ai in A) > ck = largest element in C > foreach(bj in B) > AGAIN: > if(ai + bj + ck == 0) algorithm over > if(ai + bj + ck > 0) ck change to its neighbor in C and goto AGAIN > if(ai + bj + ck < 0) continue checking next bj > > > On Sun, Nov 1, 2009 at 3:10 PM, anilkumarmyla <anilkumarm...@gmail.com> > wrote: > > No matter whatever i could think of, I am unable to do better than N^3 > > > > @daizi sheng > > I don't get your algorithm > > "2. enumerate ai, bj both in ascending order," > > Will that really help ? In what way ? > > > > -- > > > > You received this message because you are subscribed to the Google Groups > > "Algorithm Geeks" group. > > To post to this group, send email to algoge...@googlegroups.com. > > For more options, visit this group at > > http://groups.google.com/group/algogeeks?hl=en. > > > > -- > > You received this message because you are subscribed to the Google Groups > "Algorithm Geeks" group. > To post to this group, send email to algoge...@googlegroups.com. > For more options, visit this group at > http://groups.google.com/group/algogeeks?hl=en. > > > -- You received this message because you are subscribed to the Google Groups "Algorithm Geeks" group. To post to this group, send email to algoge...@googlegroups.com. For more options, visit this group at http://groups.google.com/group/algogeeks?hl=en.
