Preparation node
Fill Missing
Fill missing values with a constant or a column statistic (mean/median/mode).
What it does
Fills gaps in a column with a fixed value or a statistic computed from the column itself: mean, median, or most frequent (mode). One rule per column. You can also treat blanks or a specific sentinel as “missing,” not just nulls.
Inputs and outputs
| Inputs | One. |
| Outputs | One. |
When to use it
- Replace nulls with a sensible default
- Impute missing numbers with the mean / median
- Fill missing categories with the most common value
Examples
Constant
region → Fixed value 'Unknown'
Average
age → Mean
Numeric only; integer columns round.
Most frequent
category → Mode
Treat 'N/A' as missing
status → value 'Active', missing value 'N/A'
Under “What counts as missing.”
Tips
- Mean and median are numeric-only. Mode and a fixed value work on text, number and true/false columns; decimal, date and timestamp columns are not supported yet. For one of those, use a Formula with COALESCE([price], 0), or COALESCE([shipped], DATE '2026-01-01') for a date, or change a decimal to Double in Manage Columns first. The canvas marks such a rule before you run it.
- Statistics are computed over the whole column (a Test run uses only its sample).
- The column keeps its type, so integer statistics are rounded.