data.group

Group rows and compute aggregates.

Input: Table | Output: Table (group columns + aggregation columns)

Minimal example

Revenue by region:

nodes:
  revenue-by-region:
    type: data.group
    config:
      by: [region]
      aggregations:
        total_revenue:
          op: sum
          field: amount

Config reference

Field	Type	Required	Description
by	array	yes	Fields to group on
aggregations	map	yes	Output column name to aggregation definition

Each aggregation definition:

Field	Type	Required	Description
op	string	yes	Aggregation function
field	string	varies	Input field to aggregate (required for all except count with *)
limit	number	no	Max items for collect
separator	string	no	Delimiter for join (default: ", ")

Supported aggregates

Function	Description	Output type
count	Count of non-null values (or all rows with *)	number
sum	Sum of numeric values	number
avg	Arithmetic mean	number
min	Minimum value	same as input
max	Maximum value	same as input
first	First value in group	same as input
last	Last value in group	same as input
collect	Collect values into a JSON array	list
count_unique	Count of distinct values	number
join	Concatenate string values with separator	string

Progressive examples

Revenue by category

nodes:
  revenue-summary:
    type: data.group
    config:
      by: [category]
      aggregations:
        total_revenue:
          op: sum
          field: amount
        avg_order:
          op: avg
          field: amount
        order_count:
          op: count
          field: id

Count by status

nodes:
  status-breakdown:
    type: data.group
    config:
      by: [status]
      aggregations:
        count:
          op: count
          field: "*"

Multi-field grouping

nodes:
  region-product:
    type: data.group
    config:
      by: [region, product_type]
      aggregations:
        units_sold:
          op: sum
          field: quantity
        unique_customers:
          op: count_unique
          field: customer_id

Collect and join

nodes:
  tags-per-author:
    type: data.group
    config:
      by: [author]
      aggregations:
        all_tags:
          op: collect
          field: tag
          limit: 10
        tag_list:
          op: join
          field: tag
          separator: " | "

collect gathers values into a JSON array. limit caps the array size. join concatenates values into a single string with the specified separator.

First and last values

nodes:
  session-summary:
    type: data.group
    config:
      by: [session_id]
      aggregations:
        first_event:
          op: first
          field: event_type
        last_event:
          op: last
          field: event_type
        event_count:
          op: count
          field: "*"

Edge cases

Empty groups. If the input table is empty, the output is also empty — no group rows are produced.

NULL in group key. Rows with NULL in a group-by field are grouped together into a single group where the group key is NULL.

NULL in aggregation field. count skips NULLs (use field: "*" to count all rows including NULLs). sum and avg ignore NULL values. min and max ignore NULLs.

Single-row groups. first and last return the same value when the group has one row.

Pipeline example

name: monthly-sales-report
version: 1

nodes:
  load-sales:
    type: file.source
    path: sales.csv
    format: csv

  monthly-summary:
    type: data.group
    config:
      by: [month, region]
      aggregations:
        revenue:
          op: sum
          field: amount
        deals:
          op: count
          field: id
        top_deal:
          op: max
          field: amount

  ranked:
    type: data.sort
    config:
      by:
        - field: revenue
          direction: desc

edges:
  - "load-sales.data -> monthly-summary.input"
  - "monthly-summary.output -> ranked.input"