Find the original coordinates whose Manhattan distances are given

Given the Manhattan distances of three coordinates on a 2-D plane, the task is to find the original coordinates. Print any solution if multiple solutions are possible else print -1.
 

Input: d1 = 3, d2 = 4, d3 = 5 
Output: (0, 0), (3, 0) and (1, 3) 
Manhattan distance between (0, 0) to (3, 0) is 3, 
(3, 0) to (1, 3) is 5 and (0, 0) to (1, 3) is 4
Input: d1 = 5, d2 = 10, d3 = 12 
Output: -1 
 

Approach: Let’s analyze when no solution exists. First the triangle inequality must hold true i.e. the largest distance should not exceed the sum of other two. Second, sum of all Manhattan distances should be even. 
Here’s why, if we have three points and their x-coordinates are x1, x2 and x3 such that x1 < x2 < x3. They will contribute to the sum (x2 – x1) + (x3 – x1) + (x3 – x2) = 2 * (x3 – x1). Same logic applied for y-coordinates. 
In all the other cases, we have a solution. Let d1, d2 and d3 be the given Manhattan distances. Fix two points as (0, 0) and (d1, 0). Now since two points are fixed, we can easily find the third point as x3 = (d1 + d2 – d3) / 2 and y3 = (d2 – x3).
Below is the implementation of the above approach: 
 

# Python implementation of the approach
import math

# Function to find the original coordinated
def solve(d1, d2, d3):

    # Maximum of the given distances
    maxx = max(d1, d2, d3)

    # Sum of the given distances
    sum = (d1 + d2 + d3)

    # Conditions when the
    # solution doesn't exist
    if 2 * maxx > sum or sum % 2 == 1:
        print(-1)
        return

    # First coordinate
    x1 = 0
    y1 = 0

    # Second coordinate
    x2 = d1
    y2 = 0

    # Third coordinate
    x3 = math.floor((d1 + d2 - d3) / 2)
    y3 = math.floor((d2 + d3 - d1) / 2)
    print(f"( {x1}, {y1}), ({x2}, {y2}) and ({x3}, {y3})")

# Driver code
d1 = 3
d2 = 4
d3 = 5
solve(d1, d2, d3)

# The  code is contributed by Gautam goel (gautamgoel962)

(0, 0), (3, 0) and (1, 3)

Time Complexity: O(1)

Auxiliary Space: O(1)