¶ Description
Compute distance between points/vectors.
¶ Parameters
¶ Parameters tab
Parameters:
- Output column's name with distance
- Normalise vectors to have a 1-norm
¶ Description tab
Parameters:
- Script name
- Short description
- Revision
- Description
¶ Configuration tab
See dedicated page for more information.
¶ About
The distanceBetweenPoints action computes the distance between corresponding vectors from two input tables. It optionally normalizes the vectors using the 1-norm before computing the distance.
This tool is useful for vector similarity analysis, clustering preprocessing, and distance-based filtering operations.
¶ Input
This action requires two input tables, each containing one or more numerical columns with matching column names.
- Column name requirement: At least one column name must include an underscore (
_). This is used to identify corresponding vectors across the inputs.
¶ Example Input Tables
¶ Input Table 1
| C1 |
|---|
| 2 |
¶ Input Table 2
| C1_ |
|---|
| 4 |
Ensure that column names follow the required pattern (e.g., C1 and C1_).
¶ Output
The action returns a single table with a new column containing the calculated distances.
¶ Example Output
¶ How to Use
- Prepare two input tables with matching vector columns (e.g.,
C1andC1_). - Ensure at least one column has an underscore (
_) in its name. - Set the
Output column's name with distanceparameter to define the output column name (e.g.,Distance). - Enable or disable
Normalise vectors to have a 1-normto control 1-norm normalization before distance calculation. - Connect the two input tables to the action.
- Run the pipeline.
¶ Pipeline Example
¶ Pipeline Layout
¶ Configuration
The following default behaviors apply when converting inputs:
- Convert float to key
- If negative →
abort - If greater than 2^32-3 →
abort - If not an integer →
set to lower integer
- If negative →
- Convert string to float
- If not a number →
abort
- If not a number →
- Convert
String/?to key →abort
¶ Troubleshooting
¶ Common Errors
- Column name does not contain an underscore
error: the column 'C2' does not contains a '_' char.
✅ Solution: Rename at least one column to include an underscore, e.g., C2_.
- Column not found
error: the column 'C2' does not exists on pin 0
✅ Solution: Verify column names are identical across inputs except for the _ suffix.
- Successful execution
Success! (finished at 31/7/25 15:59:17 after 0.38 seconds)
Indicates the distance calculation ran without issues.
Notes
- This action is best suited for feature vector comparisons across datasets.
- Only supports pairwise comparison; ensure input tables have corresponding row counts.
