Tail / GroupBy Layer
Select last N rows from each group, similar to pandas groupby().tail() or dplyr's group_by() with slice_tail().
Example scenarios:
- Most Recent Customer Interactions
| Configuration | Result |
|---|---|
| by=[customer_id] | Group by customer |
| n=2 | Last 2 interactions |
| select=[date, interaction_type] | Get interaction details |
- Final Readings Per Sensor
| Configuration | Result |
|---|---|
| by=[sensor_id, date] | Group by sensor and day |
| n=3 | Last 3 readings |
| select=[timestamp, reading] | Get measurement details |
Common applications:
- Latest status updates
- Most recent transactions
- Final measurements
- Recent customer contacts
- Latest system states
Table
0
0
Table
By
[column, ...]Columns to group by. Creates separate groups for selecting last N rows. Examples:
- [machine_id]: Last N readings per machine
- [employee_id, project_id]: Last N tasks per employee per project
- [customer_id, year]: Last N purchases per customer yearly
N
u32Number of rows to select from each group (minimum 1). Common values:
- 1: Most recent item
- 3: Last three entries
- 5: Last five records
- 10: Last ten items
Select
[column, ...]Columns to include in the result. Default ['*'] selects all columns. Examples:
- [timestamp, status]: Recent states
- [transaction_id, amount]: Latest transactions
- [reading_time, value]: Final measurements