Overview
The Sorting node is a Data Preparation node in a Weave workflow that arranges rows in a specified order across one or more columns. It Sorts the records based on the chosen fields and order(asc/desc).
Sorting nodes are added from the Data Preparation category of the node panel. Configuration is one step: open the Sorting popup via the Select button, check the columns to sort by, pick each column’s direction, and drag them into priority order.
Note: “Weave” is the current name for what earlier interface versions called “Workflow.” This reference uses “Weave.”
When to use it
- Producing a specific row order for downstream nodes or destination writes that depend on order (e.g., Output nodes writing to files where row order matters).
- Multi-key sorting where ties on a primary column should be broken by secondary or tertiary columns — common in reporting (sort by Country Ascending, then by Revenue Descending within each country).
- Top-N or bottom-N analysis preparation — sort by a metric Descending, then a downstream Trimming or Filter node selects the top N rows.
- Time-series ordering — sort by a date column Ascending to ensure rows flow downstream in chronological order.
- Visual ordering for Result-Tab inspection during pipeline development.
Configuration
Sorting popup
Single configuration surface. All sorting behavior is configured in this popup.
| Element | Description |
| Select All | Checkbox at the top. Checks every upstream column for participation in the sort. |
| Per-column checkbox | Includes the column in the sort. Unchecked columns are kept in the output but don’t influence row order. |
| Per-column direction dropdown | Two options: Ascending (smallest first) or Descending (largest first). Each column has its own dropdown — directions across columns are independent and can be mixed freely. |
| Priority order (drag-and-drop) | Columns can be arranged within the popup to establish priority. The topmost column is the primary sort key; the second column breaks ties on the first; the third breaks ties on the second; and so on. |
| Apply | Commits the sort configuration and closes the popup. |
Multi-key sort semantics
When multiple columns are selected, Sorting performs a multi-key sort with the priority order set by the column order in the popup.
| Configuration | Example | Behavior |
| Single column | ID Ascending | Rows ordered by ID smallest to largest. |
| Two columns, same direction | Country Ascending, then Revenue Ascending | Rows first ordered by Country alphabetically; within each country, ordered by Revenue smallest to largest. |
| Two columns, mixed directions | Country Ascending, then Revenue Descending | Rows first ordered by Country alphabetically; within each country, ordered by Revenue largest to smallest. |
| Three columns | ID Ascending, then SEPAL_LENGTH Ascending, then PETAL_LENGTH Descending | Primary by ID; ties on ID broken by SEPAL_LENGTH Ascending; ties on SEPAL_LENGTH broken by PETAL_LENGTH Descending. |
Key behaviors
Sorting changes row order only. Unlike every other Data Preparation node documented so far — Cleansing, Imputation, Duplicates, Filter, Fuzzy Filter, Outlier Treatment, Validation, Binning, Trimming — Sorting doesn’t modify column values, doesn’t drop rows, doesn’t add columns. The output has the same rows with the same values; only the order has changed. This makes Sorting the safest Data Preparation node — it can never lose data or change values.
Per-column direction. Each selected column has its own Ascending or Descending dropdown. Directions can be mixed across columns — e.g., sort by Country Ascending and Revenue Descending within each country. This is more flexible than Trimming (one direction across one column) and unlike Duplicates Partition and Ranking (one direction across one Unique Key).
Priority arrangement. Column priority is established by the order of columns in the Sorting popup. The topmost column is the primary sort key; subsequent columns break ties in sequence. This is a UI interaction pattern unique to Sorting among Weave’s transformation nodes — most other nodes use checkboxes or dropdowns for ordering.
Multi-key sort breaks ties hierarchically. Subsequent sort keys only matter when prior keys have ties. If the primary key has all distinct values, the secondary key never acts on anything. For a deterministic order regardless of the data, ensure the last sort key has a column with no ties (e.g., a primary key) — this guarantees a stable, predictable order for every input dataset.