It is an interesting puzzle I came across , according to which , given an array , we need to find the ninja index in it.
A Ninja index is defined by these rules :
An index K such that all elements with smaller indexes have values lower or equal to A[K] and all elements with greater indexes have values greater or equal to A[K].
For example , consider :
A[0]=4, A[1]=2, A[2]=2, A[3]=3, A[4]=1, A[5]=4, A[6]=7, A[7]=8, A[8]=6, A[9]=9.
In this case, 5 is a ninja index , since A[r]<=A[5] for r = [0,k] and A[5]<=A[r] r = [k,n].
What algorithm shall we follow to find it in O(n) . I already have a brute force O(n^2) solution.
EDIT : There can be more than 1 ninja index , but we need to find the first one preferably. And in case there is no NI , then we shall return -1.
Precompute minimum values for all the suffixes of the array and maximum values for all prefixes. With this data every element can be checked for Ninja in O(1).