= Complex DB Reports (SQL, Stored Procedures, Relational Algebra) =

=== User Reputation Leaderboard ===

Report on the most valuable forgers on PCForge. The score is calculated using the following metrics:
* Productivity: Number of approved builds created (Weight: 10).
* Popularity: Number of times their builds were favourited (Weight: 5).
* Quality: The average star rating across all their builds (Weight: 20).

**SQL:**

{{{
WITH build_stats AS (
    SELECT 
        b.id AS build_id,
        b.user_id,
        COUNT(DISTINCT fb.user_id) AS favorites_count,
        AVG(rb.value) AS avg_rating
    FROM build b
    LEFT JOIN favorite_build fb ON b.id = fb.build_id
    LEFT JOIN rating_build rb ON b.id = rb.build_id
    WHERE b.is_approved = TRUE
    GROUP BY b.id, b.user_id
),
user_stats AS (
    SELECT
        user_id,
        COUNT(build_id) AS approved_builds_count,
        SUM(favorites_count) AS total_favorites_received,
        AVG(avg_rating) AS avg_rating_received
    FROM build_stats
    GROUP BY user_id
)
SELECT
    u.username,
    u.email,
    us.approved_builds_count,
    us.total_favorites_received,
    ROUND(us.avg_rating_received, 2) AS avg_rating_received,
    (
        (us.approved_builds_count * 10) +
        (us.total_favorites_received * 5) +
        (us.avg_rating_received * 20)
    ) AS reputation_score
FROM user_stats us
JOIN users u ON u.id = us.user_id
ORDER BY reputation_score DESC
LIMIT 10;
}}}

**Relational algebra:**

{{{
build_stats =
γ_{b.id → build_id, b.user_id;
   COUNT_DISTINCT(fb.user_id) → favorites_count,
   AVG(rb.value) → avg_rating
} (
  σ_{b.is_approved = TRUE} (
    (build b
      ⟕_{b.id = fb.build_id} favorite_build fb)
      ⟕_{b.id = rb.build_id} rating_build rb
  )
)

user_stats =
γ_{user_id;
COUNT(build_id) → approved_builds_count,
SUM(favorites_count) → total_favorites_received,
AVG(avg_rating) → avg_rating_received
} (
build_stats
)

λ_10 (
  τ_{reputation_score DESC} (
    π_{u.username,
      u.email,
      us.approved_builds_count,
      us.total_favorites_received,
      ROUND(us.avg_rating_received, 2) → avg_rating_received,
      ((us.approved_builds_count * 10)
       + (us.total_favorites_received * 5)
       + (us.avg_rating_received * 20)) → reputation_score
    } (
      user_stats us ⋈_{u.id = us.user_id} users u
    )
  )
)
}}}