Reputation: 53
Given 2 arrays of integers, A and B, an operation on array B is defined as follows:
B[i] = B[i]+2 and B[j] = B[j]-2, where i != j
i and j can be any indices and the above operation can be performed any number of times such that i and j are not equal
a valid operation consists of both the addition and subtraction steps, both parts are mandatory
The array is considered equal if the frequency of all the elements is same, the array need not be ordered, find the minimum operations required
Input:
A = [ 2, 10, 14 ]
B = [ 6, 2, 18 ]
Output: 2
Explanation :
1st operation: select i=0; j=2;
B[i] += 2 i.e B[0]=8;
B[j] -= 2 i.e B[2] = 16;
B after 1st operation [8,2,16]
2nd operation: select i=0; j=2;
B[i] += 2 i.e B[0]=10;
B[j] -= 2 i.e B[2] = 14;
B after 2nd operation [10,2,14]
Order doesnt matter, so we have made the arrays equal return 2;
I am unable get an approach to solve this and couldn't find any similar questions, so posting this here, thanks in advance.
Upvotes: 1
Views: 3685
Reputation: 1
A = [2, 10, 14]( % 2 == 0)
B = [2, 6, 18]( % 2 == 0)
another example
A = [1, 2, 5] -> [1, 5]( % 2 == 1) & [2]( % 2 == 0)
B = [1, 3, 4] -> [1, 3]( % 2 == 1) & [4]( % 2 == 0)
Notice that (a + k) mod k == a
.
Assuming we already have a sorted array.
We divide the array into k parts, according to the mod k value of the element, then we sum all absolute differences, it's four times the answer.
k = 2
A.sort(key=lambda x: x % k)
B.sort(key=lambda x: x % k)
result = 0
n = len(A)
for i in range(n):
result += abs(A[i] - B[i])
print(result >> 2)
# A = [1, 2, 5]
# B = [1, 3, 4]
# result = 1
# A = [2, 10, 14]
# B = [6, 2, 18]
# result = 2
O(N log N) because of sorting.
Upvotes: 0
Reputation: 9085
Assuming the arrays are solvable, then sort the arrays (by parity, and then by value), add up the absolute value of the deltas and divide by 4.
E.g.
A = [ 2, 10, 14 ], already sorted
B = [ 6, 2, 18 ], sorted becomes [2, 6, 18]
Sum of absolute value of deltas = 0 + 4 + 4 = 8. 8/4 = 2 so 2 moves.
Upvotes: 0