Module window 

Window functions for time-series query processing.

Window functions compute values over a sliding window or partition of rows without collapsing the result set (unlike aggregation, which produces fewer rows). Each input row yields exactly one output value.

Supported functions

Function	Description
RowNumber	Sequential 1-based row number within partition
Rank	Rank with gaps (1224) for ties on the order column
DenseRank	Rank without gaps (1223) for ties on the order column
Lead(offset, default)	Value offset rows ahead (default if out of range)
Lag(offset, default)	Value offset rows behind (default if out of range)
Delta	Difference between current and previous value
Rate	Per-second rate: delta(value) / delta(timestamp)
IRate	Instantaneous per-second rate between last two samples
MovingAverage(window)	Simple sliding-window mean
CumulativeSum	Running total
Increase	Monotonic counter increase (resets to 0 on counter wrap)

Enums

WindowFn

Window function type.

Functions

apply_window

Apply window functions to a RecordBatch.

window_fn_name

Human-readable function name for column naming.