A company is planning to interview 2n people. Given the array costs where costs[i] = [aCosti, bCosti], the cost of flying the i-th person to city a is aCosti, and the cost of flying the i-th person to city b is bCosti.

Return the minimum cost to fly every person to a city such that exactly n people arrive in each city.

Examples

Example 1:

Input: costs = [[10,20],[30,200],[400,50],[30,20]]
Output: 110
Explanation: 
The first person goes to city A for a cost of 10.
The second person goes to city A for a cost of 30.
The third person goes to city B for a cost of 50.
The fourth person goes to city B for a cost of 20.
The total minimum cost is 10 + 30 + 50 + 20 = 110 to have half the people interviewing in each city.

Example 2:

Input: costs = [[259,770],[448,54],[926,667],[184,139],[840,118],[577,469]]
Output: 1859

Example 3:

Input: costs = [[515,563],[451,713],[537,709],[343,819],[855,779],[457,60],[650,359],[631,42]]
Output: 3086

Constraints

  • 2 * n == costs.length
  • 2 <= costs.length <= 100
  • costs.length is even
  • 1 <= aCosti, bCosti <= 1000

Thinking Process

A company is planning to interview 2n people. Given the array costs where costs[i] = [aCosti, bCosti], the cost of flying the i-th person to city a is aCosti, and the cost of flying the i-th person to city b is bCosti.

Return the minimum cost to fly every person to a city such that exactly n people arrive in each city.

  • Greedy works when local optimal choices lead to global optimum.
  • Often sort first to make the greedy choice obvious.
  • Prove or sanity-check: would swapping two choices ever help?
Greedy choice pick locally best after sorting

Common Approaches

Typical techniques for this pattern:

Approach Time Space Notes
Sort + greedy (this problem) O(n log n) O(1) Interval scheduling, assignment
Local greedy choice O(n) O(1) Jump game, gas station
Greedy + heap O(n log n) O(n) Merge streams, room allocation
Exchange argument O(n) O(1) Prove greedy choice is safe

Solution

Solution: Greedy with Sorting by Difference

class Solution:
    def twoCitySchedCost(self, costs):
        costs.sort(key=lambda x: x[1] - x[0])

        total = 0
        n = len(costs) // 2

        for i in range(n):
            total += costs[i][1] + costs[i + n][0]

        return total

Solution Explanation

Approach: Sort + greedy (this problem)

Key idea: A company is planning to interview 2n people. Given the array costs where costs[i] = [aCosti, bCosti], the cost of flying the i-th person to city a is aCosti, and the cost of flying the i-th person to city b is bCosti.

How the code works:

  • Greedy works when local optimal choices lead to global optimum.
  • Often sort first to make the greedy choice obvious.
  • Prove or sanity-check: would swapping two choices ever help?

Walkthrough — input costs = [[10,20],[30,200],[400,50],[30,20]], expected output 110:

The first person goes to city A for a cost of 10. The second person goes to city A for a cost of 30. The third person goes to city B for a cost of 50. The fourth person goes to city B for a cost of 20. The total minimum cost is 10 + 30 + 50 + 20 = 110 to have half the people interviewing in each city.

Time: O(n log n) where n is the number of people · Space: O(1)

Algorithm Explanation:

  1. Sort by Cost Difference (Lines 4-6):
    • Sort costs by cost[0] - cost[1] in ascending order
    • Custom comparator: u[0] - u[1] < v[0] - v[1]
    • Why: People with smaller differences (cheaper to send to A) come first
    • Why: People with larger differences (cheaper to send to B) come last
  2. Initialize (Lines 7-8):
    • total: Total cost (starts at 0)
    • N: Half the number of people (costs.size() / 2)
  3. Assign and Sum (Lines 9-12):
    • For first N people (i = 0 to N-1):
      • Add costs[i][0] (send to city A)
    • For last N people (i = N to 2N-1):
      • Add costs[i + N][1] (send to city B)
    • Note: costs[i + N] accesses the (i+N)-th element, which is in the second half
  4. Return (Line 13):
    • Return the total cost

Why This Works:

Key Insight: The difference cost[0] - cost[1] tells us the “savings” from sending a person to city A instead of city B.

Strategy:

  1. Sort by difference: People with negative differences (cheaper to A) come first
  2. Send first N to A: These people save money by going to A
  3. Send last N to B: These people save money by going to B (or lose less)
  4. Optimal: This maximizes total savings

Mathematical Proof:

  • If we send person i to city A: cost = costs[i][0]
  • If we send person i to city B: cost = costs[i][1]
  • Savings from sending to A: costs[i][1] - costs[i][0] = -(costs[i][0] - costs[i][1])
  • To maximize savings, send people with largest (costs[i][1] - costs[i][0]) to A
  • Equivalently, send people with smallest (costs[i][0] - costs[i][1]) to A

Example Walkthrough:

Example 1: costs = [[10,20],[30,200],[400,50],[30,20]]

Calculate differences:

Person 0: [10, 20] → difference = 10 - 20 = -10 (cheaper to A)
Person 1: [30, 200] → difference = 30 - 200 = -170 (cheaper to A)
Person 2: [400, 50] → difference = 400 - 50 = 350 (cheaper to B)
Person 3: [30, 20] → difference = 30 - 20 = 10 (cheaper to B)

After sorting by difference (ascending):

Sorted: [[30, 200], [10, 20], [30, 20], [400, 50]]
        (diff=-170)  (diff=-10)  (diff=10)  (diff=350)

Execution:

N = 4 / 2 = 2

First N (i=0,1): Send to city A
  i=0: costs[0][0] = 30
  i=1: costs[1][0] = 10

Last N (i=0,1): Send to city B
  i=0: costs[0+2][1] = costs[2][1] = 20
  i=1: costs[1+2][1] = costs[3][1] = 50

Total = 30 + 10 + 20 + 50 = 110

Verification:

Person 0: [30, 200] → Send to A: 30 ✓
Person 1: [10, 20] → Send to A: 10 ✓
Person 2: [30, 20] → Send to B: 20 ✓
Person 3: [400, 50] → Send to B: 50 ✓
Total: 30 + 10 + 20 + 50 = 110

Example 2: costs = [[259,770],[448,54],[926,667],[184,139],[840,118],[577,469]]

Calculate differences:

Person 0: [259, 770] → diff = 259 - 770 = -511
Person 1: [448, 54] → diff = 448 - 54 = 394
Person 2: [926, 667] → diff = 926 - 667 = 259
Person 3: [184, 139] → diff = 184 - 139 = 45
Person 4: [840, 118] → diff = 840 - 118 = 722
Person 5: [577, 469] → diff = 577 - 469 = 108

After sorting by difference (ascending):

Sorted: [[259, 770], [184, 139], [577, 469], [926, 667], [448, 54], [840, 118]]
        (diff=-511)  (diff=45)   (diff=108)  (diff=259)  (diff=394)  (diff=722)

Execution:

N = 6 / 2 = 3

First N (i=0,1,2): Send to city A
  i=0: costs[0][0] = 259
  i=1: costs[1][0] = 184
  i=2: costs[2][0] = 577

Last N (i=0,1,2): Send to city B
  i=0: costs[0+3][1] = costs[3][1] = 667
  i=1: costs[1+3][1] = costs[4][1] = 54
  i=2: costs[2+3][1] = costs[5][1] = 118

Total = 259 + 184 + 577 + 667 + 54 + 118 = 1859

Algorithm Breakdown

Why Sorting by Difference Works

Key Insight: The difference cost[0] - cost[1] represents the “penalty” for choosing city A over city B.

Analysis:

  • Negative difference (cost[0] - cost[1] < 0): cost[0] < cost[1] → Cheaper to send to A
  • Positive difference (cost[0] - cost[1] > 0): cost[0] > cost[1] → Cheaper to send to B
  • Zero difference (cost[0] - cost[1] = 0): Same cost to both cities

Sorting Strategy:

  • Sort by cost[0] - cost[1] in ascending order
  • First half (smallest differences) → Send to city A (minimize penalty)
  • Second half (largest differences) → Send to city B (minimize penalty)

Why This is Optimal

Greedy Choice Property:

  • For each person, we want to minimize their cost
  • Sorting by difference ensures we make locally optimal choices
  • The constraint (exactly N people to each city) is satisfied by the split

Optimal Substructure:

  • After sorting, the problem reduces to: send first N to A, last N to B
  • This is optimal because we’re maximizing total savings

Alternative Interpretation

Cost Savings:

  • Savings from sending person i to A: costs[i][1] - costs[i][0]
  • To maximize total savings, send people with largest savings to A
  • Equivalently, send people with smallest (costs[i][0] - costs[i][1]) to A

Time & Space Complexity

  • Time Complexity: O(n log n) where n is the number of people
    • Sorting: O(n log n) - dominates the time complexity
    • Summing: O(n) - single pass through sorted array
    • Total: O(n log n)
  • Space Complexity: O(1)
    • Only using a few variables (total, N, i)
    • Sorting is typically in-place (O(1) extra space)

Key Points

  1. Sort by Difference: Sort by cost[0] - cost[1] to identify which city is cheaper
  2. Greedy Assignment: Send first N to city A, last N to city B
  3. Optimal: This greedy strategy minimizes total cost
  4. Simple: Straightforward implementation after sorting
  5. Efficient: O(n log n) time complexity

Common Mistakes

  1. All same cost: [[1,1],[2,2],[3,3],[4,4]] → Any assignment works
  2. All cheaper to A: [[1,100],[2,200],[3,300],[4,400]] → Send first 2 to A, last 2 to A (but constraint: need 2 to B)
  3. All cheaper to B: [[100,1],[200,2],[300,3],[400,4]] → Send first 2 to A, last 2 to B

  4. Mixed: [[10,20],[30,200],[400,50],[30,20]] → Example 1

  5. Wrong sorting order: Sorting in descending instead of ascending
  6. Wrong assignment: Sending first N to B instead of A
  7. Index errors: Incorrectly accessing costs[i + N]
  8. Not understanding difference: Not realizing what the difference represents
  9. Forgetting constraint: Not ensuring exactly N people to each city

Follow-Up: Why Sort by Difference?

Question: Why does sorting by cost[0] - cost[1] work?

Answer:

  • The difference tells us the “penalty” for choosing city A over city B
  • Negative difference: Choosing A saves money (penalty is negative)
  • Positive difference: Choosing A costs more (penalty is positive)
  • By sorting in ascending order, we put people who save the most by going to A first
  • We send the first N to A (maximize savings) and last N to B (minimize losses)

Mathematical Formulation:

  • Total cost if we send person i to A: costs[i][0]
  • Total cost if we send person i to B: costs[i][1]
  • Difference: costs[i][0] - costs[i][1]
  • If difference < 0: A is cheaper
  • If difference > 0: B is cheaper
  • Sorting by difference groups people by which city is cheaper for them

Tags

Array, Greedy, Sorting, Medium

Key Takeaways

  • Greedy works when local optimal choices lead to global optimum.
  • Often sort first to make the greedy choice obvious.
  • Prove or sanity-check: would swapping two choices ever help?

References

Template Reference