Skip to main content

CSV File Diff

Overview​

The CSV File Diff feature lets you compare data across CSV or other table-structured files (such as TSV, Excel, etc.). This is useful for validating file exports, ETL processes, or migrations.

Video Demonstration​

How to Use SmartDiff CSV File Diff​

Step 1: Open SmartDiff​

  • Select SmartDiff from the left panel — this opens the Workflows list

SmartDiff Workflows list

Step 2: Create a New Diff Workflow​

  • Click + Create diff. This creates a draft workflow and opens the editor directly.

Step 3: Switch to File Type​

  • In the metadata strip at the top of the editor, switch the type tabs from Data sources (default) to File
  • Source folder / Target folder pickers become available — pick a datasource to scope the file pickers, or leave blank and pick files directly per pair

Step 4: Add a File Pair​

Open the Mapping tab — for a File-type workflow this is a master-detail layout of file pairs:

  • Left rail — click Add pair to create a new pair (named file_pair_1, file_pair_2, …); each row shows its status (Files not selected / Not analyzed / Needs a key column / mapped column + key count)
  • Right pane — pick a source and target file for the pair, for example: movies_tv_data_source.csv vs movies_tv_data_target.csv

Step 5: Analyze​

  • Click Analyze once both files are picked
  • This infers the column schema and sample rows for both files, auto-maps exact-name columns into the mapping table, and collapses the file pickers to make room for the results
  • Use Re-analyze if you swap either file afterward

Click Analyze to detect columns

Step 6: Review Column Mapping​

  • The column mappings table lists each paired column with its source/target type and match status (exact, auto-fixed, name diff, AI suggested, type mismatch)
  • Toggle the key icon on a column to mark it as part of the comparison key
  • Columns that didn't get an exact-name match appear in the Unmapped columns section below — pair them manually with the dropdown + Pair button, accept high-confidence matches with Accept all fuzzy, or use Auto-map remaining

Column mappings with the key column selected

Step 7: Configure (Optional)​

  • Configuration tab — engine, summary, and storage settings
  • Validation tab — pre-flight checks; resolve anything flagged as danger before running
note

File Diff workflows don't have separate Filters or Transforms tabs — those are only available on Data Source Diff workflows. See Data Transformation for the full transform guide.

Step 8: Run and View Results​

  • Click Run diff in the header. The editor switches to the Results tab; the other tabs disable while the run is in progress.
  • The left rail lists your file pairs with status and accuracy; select All pairs for the aggregate view or a specific pair for its detail.

File Diff results

Diff Overview​

Shows high-level differences between files, in the All pairs insights view:

  • Rows compared, Matches, Diffs, Accuracy stat cards
  • Root cause clusters and a per-pair accuracy bar list

Summary​

  • All data: All columns with counts
  • Only diff: Only differing columns
  • By source / By target: Frequency-based comparisons
  • Export: CSV/PDF

Meta Diff​

  • Shows metadata differences per column (name, datatype, etc.)
  • Highlights differences with color coding
  • Export available from the chrome row

Data Diff​

  • Groups differences into clusters
  • Only diffs toggle plus row search
  • Hide non-relevant columns
  • Export: CSV/PDF