Pivot Longer

Reshape

Stack multiple columns into key-value rows, converting wide-format data into long-format so that each observation occupies its own row. Long format is the standard layout expected by most statistical modelling functions, mixed-effects models, and many plotting libraries.

When to Use

Each measurement lives in a separate column (e.g., Day1, Day2, Day3) and you need them stacked into a single column for analysis. Most repeated-measures and mixed-effects models require long format.
You are preparing data for repeated-measures tests, mixed-effects models, or plotting tools (like ggplot2 or seaborn) that expect one observation per row.
Column headers contain variable values (time points, doses, conditions) that should be recorded as data values in their own column, not as structural column names.
You need to combine several measurement columns into a single column to calculate summary statistics across all measurements uniformly.

Required Inputs

Columns to Pivot -- the columns to stack into rows (e.g., Day1, Day2, Day3). Select all the columns that represent the same type of measurement taken under different conditions.
Names To -- the name for the new column that will receive the original column headers (e.g., "Day" or "TimePoint"). This column records which original column each value came from.
Values To -- the name for the new column that will receive the cell values (e.g., "Measurement" or "Score"). This column holds the actual data values.

What Changes

Row count increases because each pivoted column generates one new row per original row. If you pivot 3 columns, each original row becomes 3 rows in the result.
Column count decreases because all pivoted columns merge into exactly two new columns: one for the names and one for the values.
Non-pivoted columns (ID columns such as Subject or PatientID) are repeated for each new row to maintain the relationship between observations and their source entity.
The total number of data values is preserved -- the same information is simply rearranged from a wide layout into a tall layout.

Common Pitfalls

Pivoting columns with mixed data types (numeric and text) can cause unexpected string conversions in the values column. Ensure all columns being pivoted contain the same type of data.
Forgetting to exclude ID columns from the pivot selection will stack IDs alongside measurements, corrupting the data structure. Only select the measurement columns you want to stack.
Very wide datasets with many columns can produce a very large number of rows after pivoting (original rows x pivoted columns), potentially slowing the interface for very large files.
If the original wide-format data has missing values in some columns, these become explicit missing value rows in the long format. Decide whether to keep or drop them after pivoting.

Example: Before & After

Before

Subject	Pre	Post
S1	80	92
S2	75	88

→

After

Subject	Time	Score
S1	Pre	80
S1	Post	92
S2	Pre	75
S2	Post	88

Columns to Pivot: Pre, Post. Names To: Time, Values To: Score. Each subject's single row expands into two rows, one per time point.

Pivot Wider

Advanced Filter