name: op-agg-min
description: Smallest non-null value of the field.
kind: operator
category: AGG
operator: AGG_MIN
type: reference
applies_to: process, compose, predict
examples_tags: [financial, streaming-friendly]
Use when
Smallest value of a numeric or date field, over all rows or per group.
Questions it answers:
- What is the earliest order date in each region?
- Are there negative or impossible values in the age field?
Use something else:
AGG_MAXwhen you want the largest value.AGG_PERCENTILEwhen one extreme row would mislead and you want a low value that ignores it.
Params
None.
Inputs
| Param | Accepted field types |
|---|---|
Field | numeric: u8/u16/u32/u64, f32/f64, decimal128, date, datetime, packed_bool, u4 |
Output
Scalar float64. Per-group when wired under a grouper.
Components
Universal floor {n, n_null} plus operator-specific:
| Key | Type | Notes |
|---|---|---|
min | float64 | Smallest non-null value observed |
- Mergeability:
Mergeable - Streaming: per-chunk emits running min
Gotchas
- Null rows skipped;
n_nullcounts them. - All-null cohort → emits NaN (sentinel for “no data”).
- Decimal128 preserved on input; float64 output coerces the comparison space.
See
pulse_examples_search tags=[streaming-friendly]- Skills:
aggregation-design,response-components