Watch the video solution for Accepted Candidates From the Interviews, a medium level problem involving Database. This walkthrough by Everyday Data Science has 2,974 views views. Want to try solving it yourself? Practice on FleetCode or read the detailed text solution.
Table: Candidates
+--------------+----------+ | Column Name | Type | +--------------+----------+ | candidate_id | int | | name | varchar | | years_of_exp | int | | interview_id | int | +--------------+----------+ candidate_id is the primary key (column with unique values) for this table. Each row of this table indicates the name of a candidate, their number of years of experience, and their interview ID.
Table: Rounds
+--------------+------+ | Column Name | Type | +--------------+------+ | interview_id | int | | round_id | int | | score | int | +--------------+------+ (interview_id, round_id) is the primary key (combination of columns with unique values) for this table. Each row of this table indicates the score of one round of an interview.
Write a solution to report the IDs of the candidates who have at least two years of experience and the sum of the score of their interview rounds is strictly greater than 15.
Return the result table in any order.
The result format is in the following example.
Example 1:
Input: Candidates table: +--------------+---------+--------------+--------------+ | candidate_id | name | years_of_exp | interview_id | +--------------+---------+--------------+--------------+ | 11 | Atticus | 1 | 101 | | 9 | Ruben | 6 | 104 | | 6 | Aliza | 10 | 109 | | 8 | Alfredo | 0 | 107 | +--------------+---------+--------------+--------------+ Rounds table: +--------------+----------+-------+ | interview_id | round_id | score | +--------------+----------+-------+ | 109 | 3 | 4 | | 101 | 2 | 8 | | 109 | 4 | 1 | | 107 | 1 | 3 | | 104 | 3 | 6 | | 109 | 1 | 4 | | 104 | 4 | 7 | | 104 | 1 | 2 | | 109 | 2 | 1 | | 104 | 2 | 7 | | 107 | 2 | 3 | | 101 | 1 | 8 | +--------------+----------+-------+ Output: +--------------+ | candidate_id | +--------------+ | 9 | +--------------+ Explanation: - Candidate 11: The total score is 16, and they have one year of experience. We do not include them in the result table because of their years of experience. - Candidate 9: The total score is 22, and they have six years of experience. We include them in the result table. - Candidate 6: The total score is 10, and they have ten years of experience. We do not include them in the result table because the score is not good enough. - Candidate 8: The total score is 6, and they have zero years of experience. We do not include them in the result table because of their years of experience and the score.
Problem Overview: You need to identify candidates who successfully pass the interview process. Each candidate participates in multiple interview rounds with recorded scores. A candidate is considered accepted only if they pass every required round, which requires joining interview data and validating scores across all rounds.
Approach 1: Join Tables + Grouping + Filtering (O(n log n) time, O(n) space)
The key idea is to combine candidate information with their interview results and then evaluate performance across all rounds. Start by joining the candidate table with the interview results table using the candidate identifier. This produces a dataset containing every candidate and their scores for each round. SQL JOIN operations allow you to merge these records efficiently before aggregation.
Next, aggregate results by candidate using GROUP BY candidate_id. Aggregation enables you to evaluate performance across multiple rounds for the same candidate. Functions such as COUNT() determine how many rounds a candidate attended, while MIN(score) or similar checks ensure the candidate passed every round. If the lowest score across all rounds still meets the passing threshold, the candidate did not fail any stage.
Filtering happens in the HAVING clause. Unlike WHERE, HAVING works on aggregated results. You can enforce rules such as passing every round or completing all required rounds by combining conditions like MIN(score) and COUNT(). This step eliminates candidates who failed at least one interview stage.
This approach relies heavily on relational operations: table joins, grouping, and aggregated filtering. These patterns appear frequently in database and SQL interview questions. The same logic can be replicated in Pandas using merge(), groupby(), and aggregation filters, which mirrors how relational queries operate in analytical workflows.
Recommended for interviews: Interviewers expect a clean relational query that joins the tables, groups by candidate, and filters with HAVING. A brute-force row-by-row check would show basic understanding, but the grouped SQL solution demonstrates strong command of relational aggregation and is the standard approach for database interview problems.
| Approach | Time | Space | When to Use |
|---|---|---|---|
| Join + Group By + HAVING | O(n log n) | O(n) | Standard relational query when validating conditions across multiple interview rounds |
| Join + Aggregation in Pandas | O(n log n) | O(n) | Data analysis workflows where interview data is processed in Python using DataFrames |
