Skip to main content

​
What This Node Does

The Union node stacks datasets vertically by appending rows from multiple sources. Use it to combine data with the same or similar column structure—like merging sales data from different regions or combining historical and current data. [SCREENSHOT: Union node with two input connections showing “Combined 10,000 rows (Region A) + 8,000 rows (Region B) = 18,000 rows”]

​
When to Use This Node

Use the Union node when you need to:
  • Combine regional data - Merge US sales + UK sales + EU sales into one dataset
  • Merge time periods - Combine Q1 data + Q2 data + Q3 data + Q4 data
  • Consolidate sources - Stack data from multiple CSV files with same structure
  • Append new to historical - Add today’s transactions to historical transaction table

​
Step-by-Step Usage Guide

1

Add Union node to canvas

2

Connect first input

Connect your first dataset to the Union node’s top input handle[SCREENSHOT: Region A data connected to top input]
3

Connect second input

Connect your second dataset to the Union node’s bottom input handle[SCREENSHOT: Region B data connected to bottom input]
4

Choose union type

Select UNION ALL (keep all rows including duplicates) or UNION (remove duplicates)[SCREENSHOT: Union type dropdown]
5

Configure column alignment

Choose “Align by Name” (default) or “Align by Position”[SCREENSHOT: Alignment options]
6

Preview combined results

​
Tips and Best Practices

UNION ALL for Performance: Use UNION ALL unless you specifically need deduplication. UNION ALL is much faster.
Add Source Column: Before union, add column identifying source (Add node: source = "Region A"). Makes it easy to filter by source later.
Match Schemas: Ensure column names match across inputs. Use Rename node before Union to standardize names.
Filter Before Union: Reduce rows in each input before combining. More efficient than filtering combined result.
Type Consistency: Convert column types before Union if needed. Text + Integer for same column creates issues.
Order Matters: Inputs stacked in connection order (top to bottom). Connect in logical order (chronological, regional, etc.).