Pulls a single part out of a date or time — the same as EXTRACT, but as a regular function.
DATE_PART('<field>', <date or time>)
Returns the year, month, day, hour and so on. It does the same as EXTRACT, except the field is written in quotes as the first argument: DATE_PART('month', created_at). The difference is the type: DATE_PART returns double precision, while EXTRACT returns numeric.
orders
| id | created_at | amount |
|---|---|---|
| 1 | 2026-07-03 10:15:00 | 120.50 |
| 2 | 2026-07-28 18:40:00 | 80.25 |
| 3 | 2026-08-09 09:05:00 | 199.99 |
| 4 | 2026-08-21 14:30:00 | 45.00 |
| 5 | 2026-09-02 11:00:00 | 64.75 |
These tables already exist in the sandbox — the examples below use them. Want to try this in your own database? Copy the SQL.
| year | hour |
|---|---|
| 2026 | 15 |
| id | created_at | month |
|---|---|---|
| 1 | 2026-07-03 10:15:00 | 7 |
| 2 | 2026-07-28 18:40:00 | 7 |
| 3 | 2026-08-09 09:05:00 | 8 |
| 4 | 2026-08-21 14:30:00 | 8 |
| 5 | 2026-09-02 11:00:00 | 9 |
| seconds |
|---|
| 7200 |
For an interval, 'epoch' is its length in seconds.
The field goes in quotes: DATE_PART(month, created_at) without quotes throws column "month" does not exist, because PostgreSQL looks for a column.
It returns a number, not a name: DATE_PART('month', …) gives 9, not September. The name comes from TO_CHAR(created_at, 'FMMonth').
Like EXTRACT, it's no good for grouping by month: September of different years merges together. Use DATE_TRUNC.