Recipe Stages Reference

Complete reference for every recipe stage — Filter, Join, Aggregate, Pivot, Impute, Sort, Sample, and Custom SQL

StageDescriptionInputOutput
FilterSelect rows matching conditions. Supports AND/OR logic, comparison operators, and NULL checks.Any datasetFiltered dataset
JoinCombine two datasets on matching columns. INNER, LEFT, RIGHT joins supported.Two datasetsJoined dataset
AggregateGroup by columns and compute summary statistics: COUNT, SUM, AVG, MIN, MAX, and custom expressions.Any datasetAggregated dataset
PivotReshape data — convert rows to columns. Specify index, columns, and values.Any datasetPivoted dataset
ImputeFill missing values using mean, median, mode, constant, forward-fill, or backward-fill strategies.Any datasetDataset with filled nulls
SortOrder rows by one or more columns, ascending or descending.Any datasetSorted dataset
SampleSelect a random subset — by count or percentage. Supports seed for reproducibility.Any datasetSampled dataset
Custom SQLWrite arbitrary SQL against the current dataset. Full power of the target backend.Any datasetSQL result dataset
Stages are chained in sequence. The output of each stage becomes the input to the next. The SQL Execution Engine compiles the entire pipeline into a single optimised query where possible, or splits it at capability boundaries.
← Recipes 📋 Contents Dataset Transforms →