I need to check for duplicate and near-duplicate rows in the Zomato Restaurant Dataset. Let me start by examining exact duplicates based on the key identifier, then look for near-duplicates using business-relevant columns. ```json { "action": "tool_call", "reasoning": "First, I'll check for exact duplicates by looking at Restaurant ID (the grain identifier) and also check for duplicate combinations of Restaurant Name + Address + City, which would indicate the same physical restaurant listed multiple times.", "tool": "query_sql", "args": { "sql": "WITH duplicate_ids AS (SELECT \"Restaurant ID\", COUNT(*) as id_count FROM raw.\"Zomato Restaurant Dataset\" GROUP BY \"Restaurant ID\" HAVING COUNT(*) > 1), duplicate_locations AS (SELECT \"Restaurant Name\", \"Address\", \"City\", COUNT(*) as location_count FROM raw.\"Zomato Restaurant Dataset\" GROUP BY \"Restaurant Name\", \"Address\", \"City\" HAVING COUNT(*) > 1) SELECT (SELECT COUNT(*) FROM duplicate_ids) as restaurants_with_duplicate_ids, (SELECT COUNT(*) FROM duplicate_locations) as duplicate_name_address_city_combos, (SELECT SUM(id_count) FROM duplicate_ids) as total_rows_with_duplicate_ids, (SELECT SUM(location_count) FROM duplicate_locations) as total_rows_with_duplicate_locations", "purpose": "Check for exact duplicates by Restaurant ID and by Restaurant Name + Address + City combination" } } ```