Page tree


You can import one or more file-based datasets from your file datastore into the Library or add them to a flow.

NOTE: When you import a file, the data is not stored in Trifacta®. What you create is an imported dataset, which is simply a reference to the source of the data.


  • You must have read permissions on any directory and file that you wish to import.


  1. From the menubar, click Library.
  2. In the Library page, click Import Data.
  3. In the left nav bar of the Import Data page, you can select the connection for the backend datastore where your file or files are located. 

    Tip: If you do not see your datastore in the left nav bar, you can create a connection to enable Trifacta to use data from it. See Create a Connection.

  4. In the Import Data page, select the required connection where the data is stored.

  5. Navigate to the required folder. 

    1. To import all files in the folder, click the plus icon next to the folder.
    2. In the folder, you can click the plus icon next to one or more files.  

      Tip: You can select multiple files from the folder for import.

      Tip: When browsing for data on the storage, you can choose to parameterize elements of the path.

  6. In the right panel, click Edit Settings for your file(s) to modify settings applied to how individual files is imported.

    NOTE: By default, the application applies a few steps to file-based imported datasets to attempt to organize them into tabular format and hides these steps from your recipe. As needed, you can disable these automated steps, so that the steps themselves appear in the Recipe panel. Deselect the Detect Structure checkbox.

    NOTE: If your file uses a different file encoding than the default encoding, you can change it for the file during the import process.

  7. Click Continue to add the dataset to the flow.

This page has no comments.