Given an infinite number line from -INFINITY to +INFINITY and we are on zero. We can move n steps either side at each n’th time.
Approach 1 : Using Tree
1st time; we can move only 1 step to both ways, means -1 1; 2nd time we can move 2 steps from -1 and 1; -1 : -3 (-1-2) 1(-1+2) 1 : -1 ( 1-2) 3(1+2) 3rd time we can move 3 steps either way from -3, 1, -1, 3 -3: -6(-3-3) 0(-3+3) 1: -2(1-3) 4(1+3) -1: -4(-1-3) 2(-1+3) 3: 0(0-3) 6(3+3) Find the minimum number of steps to reach a given number n.
Input : n = 10 Output : 4 We can reach 10 in 4 steps, 1, 3, 6, 10 Input : n = 13 Output : 5 We can reach 10 in 4 steps, -1, 2, 5, 9, 14
This problem can be modeled as tree. We put initial point 0 at root, 1 and -1 as children of root. Next level contains values at distance 2 and so on.
0 / \ -1 1 / \ / \ 1 -3 -1 3 / \ / \ / \ / \
The problem is now to find the closes node to root with value n. The idea is to do Level Order Traversal of tree to find the closest node. Note that using DFS for closest node is never a good idea (we may end up going down many unnecessary levels).
Below is C++ implementation of above idea.
Found number n at level 5
The above solution is contributed by Mu Ven.
Approach 2 : Using Vector
The above solution uses binary tree for nth time instance i.e. -n and n. But as the level of tree increases this becomes inefficient. For values like abs(200) or more above program gives segmentation fault.
Below solution does not make a tree and takes complexity equal to exact number of steps required. Also the steps required are printed in the array which equals the exact sum required.
- Distance between +n and -n is 2*n. So if you negate a number from +ve to -ve it will create difference of 2*n from previous sum.
- If a number lies between n(n+1)/2 and (n+1)(n+2)/2 for any n then we go to step (n+1)(n+2)/2 and try to decrease the sum to the difference using idea discussed above.
- If we go to n(n+1)/2 and then try to increase than it will ultimately lead you to same number of steps.
And since you cannot negate any number (as sum is already less than required sum) from n(n+1)/2 this proves that it takes minimum number of steps.
Minimum number of Steps: 7 Step sequence: 1 2 3 -4 5 6 7
If n is the sum that it is required and s is the minimum steps then:
n = (s+1)*(s+2)/2 + 1 (or +2)
Hence n = O(s*s)
Therefore s = O(sqrt(n))
Space Complexity : O(sqrt(n))
Time complexity : O(sqrt(n))
Please write comments if you find anything incorrect, or you want to share more information about the topic discussed above.
- Find the node with minimum value in a Binary Search Tree
- Write a Program to Find the Maximum Depth or Height of a Tree
- Root to leaf path sum equal to a given number
- Find the largest BST subtree in a given Binary Tree | Set 1
- Applications of Minimum Spanning Tree Problem
- Find the maximum sum leaf to root path in a Binary Tree
- Segment Tree | Set 2 (Range Minimum Query)
- Find multiplication of sums of data of leaves at same levels
- Iterative Method to find Height of Binary Tree
- Find depth of the deepest odd level leaf node
- Find next right node of a given key
- Find LCA in Binary Tree using RMQ
- Find distance between two nodes of a Binary Tree
- Find the maximum path sum between two leaves of a binary tree
- Minimum no. of iterations to pass information to all nodes in the tree
Improved By : Shreyans Vora