Pivot Table Profit Analyzer

An analytics platform stores order data in a warehouse where each record is a row of strings. The first row holds the column headers; every row after that is an order entry with string values. At minimum, the following columns exist in each record: order_id, cost, sell_price, product, and date (written as "YYYY-MM-DD"). Extra columns (for instance state) may appear as well.

Build a DataWarehouse class that lets you query a pivot-style summary: compute the net profit (sell_price − cost) for every distinct value of a chosen column (pivotColumn), considering only orders whose date is on or after a given startDate. The pivotColumn could be "state", "product", or any other column present in the header.

The class must support:

class DataWarehouse {
    public DataWarehouse(List<List<String>> data)
    public String getMostProfitable(String pivotColumn, String startDate)
}

• Constructor – Receives the complete dataset. The first inner list is the header; the remainder are the data rows.
• getMostProfitable – Looks at records with date >= startDate. For each distinct value in pivotColumn, totals the net profit and returns the value with the largest total.
  • If two values tie, return the lexicographically smallest string.
  • If no records satisfy the date filter, return "".

Constraints

• data always has at least the header row.
• Every row has the same number of entries as the header.
• All values are non‑null strings.
• Dates use the "YYYY-MM-DD" format.
• 1 ≤ data.size() ≤ 10^5
• 1 ≤ data[0].size() ≤ 20

Examples

Example

Input:
["DataWarehouse", "getMostProfitable", "getMostProfitable", "getMostProfitable", "getMostProfitable"]

[
  [["order_id", "cost", "sell_price", "product", "date", "state"],
   ["23", "12", "18", "cheese", "2023-12-04", "CA"],
   ["24", "5", "12", "melon", "2023-12-04", "OR"],
   ["25", "25", "31", "cheese", "2023-12-05", "OR"],
   ["26", "4", "12", "bread", "2023-12-05", "CA"],
   ["25", "10", "14", "cheese", "2023-12-06", "CA"],
   ["26", "5", "6", "bread", "2023-12-06", "OR"]],
  ["state", "2023-12-05"],
  ["product", "2023-12-05"],
  ["product", "2025-12-05"],
  ["color", "2023-12-01"]
]

Output:
[null, "CA", "cheese", "", ""]

Explanation

• DataWarehouse warehouse = new DataWarehouse(data);
• warehouse.getMostProfitable("state", "2023-12-05"); → "CA". Rows on or after 2023-12-05: CA net profit = (12−4) + (14−10) = 8 + 4 = 12; OR net profit = (31−25) + (6−5) = 6 + 1 = 7.
• warehouse.getMostProfitable("product", "2023-12-05"); → "cheese". Profits after that date: cheese = 10, bread = 9, melon = 0.
• warehouse.getMostProfitable("product", "2025-12-05"); → "", because no date qualifies.
• warehouse.getMostProfitable("color", "2023-12-01"); → "", because the column "color" does not exist in the header.