pivot

pivot turns tidy/long data into a wide summary table. Index columns stay as row identity; distinct values from one column become output columns.

It is useful for crosstabs and compact comparison tables. It is also one of the easiest nodes to make unreadable if the pivot column has too many distinct values.

Pivot contract

Field	Required	Notes
inputs	yes	Exactly one upstream table.
index	yes	Columns that remain as row identity in the wide result.
columns	yes	Categorical column whose distinct values become output columns.
values	yes	Numeric value column to aggregate into each pivot cell.
agg	no	sum by default; also supports mean and count.

What each cell means

• values supplies the numeric value for each cell.
• agg decides how multiple rows in the same bucket collapse. The default is sum; use mean or count when that is the actual question.
• Empty buckets become null for sum and mean; count returns counts.

Reviewing the result

default is the pivoted wide table. Inspect total width and the generated column names before sending it to reports or scripts.

Example

- id: monthly
  kind: pivot
  inputs: [sales]
  index: [region]
  columns: month
  values: revenue
  agg: sum                        # sum | mean | count, default sum

Related

• aggregate - grouped summaries without widening