Join and transform: turn two messy exports into one clean table
Bring two data sources onto one canvas
Real analysis rarely starts from a single tidy file. KNIME ships 300+ connectors, so you can read an Excel sheet, a CSV, a database table, or a cloud source and bring them together on one canvas. The first move in any join-and-transform pipeline is simply getting both inputs in.
- 1 Add two reader nodes to the canvas — for example a CSV Reader and an Excel Reader (or a second CSV Reader) — each pointed at a different export from your work.
- 2 Execute both readers and confirm each shows a green light and a populated output table.
- 3 Inspect both tables and note a column they share (a key you could join or stack on).
- 4 Decide whether you want to stack the two tables (Concatenate) or rename columns first so they line up (Column Renamer).
Two reader nodes are green on one canvas, and you have identified how their tables relate.
Concatenate vs Column Renamer — combining and tidying as nodes
Combining data in KNIME is itself just more nodes. A Concatenate node stacks tables together; a Column Renamer tidies column names so downstream nodes (and your collaborators) can make sense of them. Because each transform is a visible node, anyone opening the workflow can see exactly how the messy inputs became the clean table.
Cleaning and combining data in KNIME is done with ordinary nodes — Concatenate to stack tables, Column Renamer to tidy names — wired in the same drag-and-drop way as everything else. The transform logic is never hidden in a script, so the path from raw instrument exports to one clean, analysis-ready table stays fully visible and reproducible. This is exactly the no-code recipe you can hand to a collaborator who has no Python environment.
- ?When would you Concatenate two tables versus rename columns so they align first?
- ?If a collaborator questions one value in your clean table, how does the node canvas let you trace where it came from?
- ?What makes this combine-and-tidy step reproducible the next time the instrument exports fresh data?
Produce one clean, analysis-ready table
Now wire the combine and tidy steps into a single short pipeline that turns two messy exports into one table you would actually analyse.
Combine your two reader outputs into a single clean table using Concatenate and Column Renamer.
- 1 Wire both readers into a Concatenate node to stack their rows into one table.
- 2 Add a Column Renamer (before or after the Concatenate, as needed) to give the combined table tidy, consistent column names.
- 3 Optionally add a Row Filter to drop rows you do not want.
- 4 Execute the chain to all-green and open the final table to confirm it is clean and analysis-ready.
- 5 Save the workflow as a reusable recipe.
A saved KNIME workflow that reads two sources and outputs one clean, correctly named table.