Write a PostgreSQL query that cleans inconsistent department, category, and amount fields from CU Denver Finance expense staging data, then returns department-level totals for valid expenses only. Exclude unmatched departments and invalid or negative amounts, and keep only departments with at least 2 valid cleaned rows.
| Column | Type | Description |
|---|---|---|
| expense_id | INT | Raw expense row ID |
| dept_name_raw | VARCHAR(100) | Department name from source extract |
| category_raw | VARCHAR(50) | Raw category text |
| amount_raw | VARCHAR(30) | Amount stored as text |
| expense_date | DATE | Expense date |
| Column | Type | Description |
|---|---|---|
| dept_code | VARCHAR(10) | Department code |
| dept_name_standard | VARCHAR(100) | Standard department name |
| FIN | Finance |
| BUD | Budget Office |
| SA | Student Affairs |
| 1 | finance | travel | $1,200.50 | 2024-01-10 |
| 4 | Budget Office | supplies | -50 | 2024-01-18 |
| 8 | Student Affairs | travel | abc | 2024-02-01 |
| 10 | Unknown Dept | Supplies | $100 | 2024-02-05 |