Working with versions


All data within a dataset is encapsulated in discrete, immutable versions. The version interface allows you import data from various sources and file formats as well as clean imported and/or existing data.

When creating the first version of a dataset, you must always first import data from some other source. When working with subsequent version, you have the option of either importing data and/or cleaning or otherwise modifying the existing data.

Subsequent versions will always build on the previous version. However, you may choose to completely overwrite the existing content by replacing your current version with new uploaded data.

No changes will be made to your live dataset until you release your finalized version. You can feel free to upload and remove test files without having any affect on others working with your data.

Only users with edit access to a dataset will be able to view and use unreleased versions.


The Edits tab of the version interface allows you to upload and clean data. You can learn more about available edits in the cleaning data section.


The Table tab of the version interface shows the current table for the version, with all valid edits automatically applied. This table contains the same functionality as other tables across Redivis, though note that the cell preview is not available until the version has been released.

You can also edit metadata (variable labels, descriptions, and value labels) from this tab.


The Release tab of the version interface allows you to provide additional documentation and metadata about the version, as well as choose sampling information, when applicable. Learn more in the version release documentation.

Storage and archival

Redivis computes row-level diffs for each version, efficiently storing the complete version history in one master table. This allows you to regularly release new versions and maintain a robust version history without ballooning storage costs.

After a time, the tables associated with historic versions will automatically be archived. Archived tables can still be used and queried just like any other table, with the sole limitation being that the cell preview is not available on archived tables.