Greatest order to maximise the worth

Given, an array arr[] of N integers, the duty is to pick out some integers from this array arr[] and prepare them in one other array brr[], such that the sum of brr[i]*(i+1) for each index in array brr is most. Discover out the utmost potential such worth for a given array arr.
Enter: N = 5, arr[] = {-1, -8, 0, 5, -9}
Output: 14
Clarification: By deciding on -1, 0, and 5 we’ll get the utmost potential worth.(-1*1 + 0*2 + 5*3 = 14)Enter: N = 3, arr[] = {-1, -8, -9}
Output: 0
Clarification: Since all of the numbers are unfavorable it’s higher to not choose any quantity.
Strategy: To resolve the issue observe the under thought:
The issue may be solved utilizing a grasping strategy. The instinct behind the grasping strategy is that the utmost worth is obtained by deciding on the biggest quantity first, and the smallest quantity final. By sorting the array in lowering order, we will be sure that we have now the biggest quantity at first, and thus maximize the worth.
Steps that have been to observe the above strategy:
- First, we kind the enter array in lowering order.
- Then, we iterate over the array and calculate the prefix sums of the weather, including every prefix sum to the consequence so long as it stays non-negative.
- If the prefix sum turns into unfavorable at any level, we cease the iteration and return the present consequence.
- By stopping the iteration as quickly because the prefix sum turns into unfavorable, we will be sure that we don’t choose any numbers which have a web unfavorable contribution to the full worth. This enables us to acquire the utmost worth.
Beneath is the code to implement the above strategy:
Python3
|
Time Complexity: O(N*logN), the place N is the size of the array.
Auxiliary Area: O(1), as we aren’t utilizing any further area.