Unpivot — unpivot
Reshape wide → long (pandas melt).
Use cases
- Turn
Jan/Feb/… columns back intomonth+valuerows. - Normalize a wide spreadsheet into a tidy, tall format for grouping.
What it does
Unpivot is the inverse of Pivot: it collapses multiple columns into two columns
— one naming the original column (var_name) and one holding the value
(value_name).
Before
| region | Jan | Feb |
|---|---|---|
| North | 100 | 150 |
| South | 80 | 200 |
2 rows · 3 cols
After
| region | monthnew | amountnew |
|---|---|---|
| North | Jan | 100 |
| North | Feb | 150 |
| South | Jan | 80 |
| South | Feb | 200 |
4 rows · 3 cols
Configuration
| Config key | Type | Required | Description |
|---|---|---|---|
id_vars | string[] | Yes | Columns to keep as identifiers |
value_vars | string[] | No | Columns to unpivot (defaults to the rest) |
var_name | string | No | Name for the variable column (default variable) |
value_name | string | No | Name for the value column (default value) |
Generated Python code
These exports keep region and unpivot every other column using the default
output names variable and value. Null values are retained. The illustration
above instead names these columns month and amount.
pandas
python
df_2 = df_1.melt(id_vars=['region'])Polars
python
df_2 = df_1.unpivot(index=['region'])Tips & common mistakes
- Leave
value_varsempty to unpivot every column that isn't anid_var. - Name
var_name/value_namemeaningfully (e.g.month/amount) for a tidy result. - To go the other way (long → wide), use Pivot.