check-data-against-backup

To compare two CSV files for cell-level differences, provide the current and reference file paths to generate a structured diff report highlighting exact value changes with float tolerance, schema changes, and row count checks.

Yes, you can validate dataset changes against a DVC cache backup by providing a .dvc file path, which the tool resolves to the cached reference version for cell-level diffing and schema validation.

Supported formats for cell-level data diffing include CSV, TSV, Parquet, and Excel, allowing you to compare current data files against prior versions or backups across these common dataset formats.

To verify a data pipeline only changed expected cells, compare the updated dataset against a reference backup to produce a structured verdict highlighting schema changes, row counts, and specific cell-level diffs with float tolerance.

Yes, the dataset comparison tool handles floating point tolerance in diffs, allowing minor numerical variations to pass validation while still reporting significant cell-level value changes accurately.

The best way to detect schema and row count changes between dataset versions is to run a cell-level diff against a reference backup, which outputs a structured verdict identifying structural changes and exact value modifications.