Compute the Rank as a Percentage — LeetCode 2346 Python Solution

Table: Students +---------------+------+ | Column Name | Type | +---------------+------+ | student_id | int | | department_id | int | | mark | int | +---------------+------+ student_id contains unique values. Each row of this table indicates a student's ID, the ID of the department in which the student enrolled, and their mark in the exam.

+---------------+------+
| Column Name   | Type |
+---------------+------+
| student_id    | int  |
| department_id | int  |
| mark          | int  |
+---------------+------+
student_id contains unique values.
Each row of this table indicates a student's ID, the ID of the department in which the student enrolled, and their mark in the exam.

import duckdb
import pandas as pd

def solution(students: pd.DataFrame) -> pd.DataFrame:
    con = duckdb.connect()
    con.register("Students", students)
    return con.execute("""SELECT
    student_id,
    department_id,
    IFNULL(
        ROUND(
            (
                RANK() OVER (
                    PARTITION BY department_id
                    ORDER BY mark DESC
                ) - 1
            ) * 100 / (COUNT(1) OVER (PARTITION BY department_id) - 1),
            2
        ),
        0
    ) AS percentage
FROM Students;""").df()