Transpose
Unpivot wide columns into tall Name/Value rows (the inverse of Cross Tab).
What it does
Unpivots wide data into tall data, the inverse of Cross Tab. Pick the Key fields (identity columns that stay); the remaining Data columns collapse into two new columns: Name (the original column's name) and Value (that cell's value). Every input row becomes one row per data column. The Value column keeps the common type of the data columns (all numbers stay numeric; mixed types fall back to text), and because the output columns are known from your settings, they appear in the builder without a test run.
Inputs and outputs
| Inputs | One. |
| Outputs | One. |
When to use it
- Normalize a wide spreadsheet export (months or categories spread across columns) into tidy long form for a database or BI tool
- Reshape one-column-per-metric data into one metric per row before grouping, charting, or joining
- Undo a Cross Tab, or build a column-agnostic pipeline over an unknown-width table
Examples
Months to rows
Key: id · Data: all other columns
id, jan, feb, mar → one row per id per month (Name = jan/feb/mar, Value = the amount).
Specific columns only
Key: id · Choose columns: q1, q2
Melt just the two quarters; leave the rest wide.
Drop the blanks
Turn on “Drop null values”
Skip melted rows whose value is empty, with no downstream Filter needed.
Tips
- “All other columns” melts everything that isn’t a key, and new upstream columns are included automatically. Switch to “Choose specific columns” to pin an exact set.
- Value keeps its natural type when the data columns agree (numbers stay numbers); set Value type to Text to force a string.
- Rename the Name and Value columns to whatever fits (e.g. month / amount).
- Add a Position column to record each field’s order (jan=1, feb=2, mar=3).
- It streams, so memory stays bounded even on very large inputs.