Load

Load lets you load data from several sources into your current session. Specifically, you can load data from:

• Datasets
• Files (legacy)
• Snapshots (legacy)
• URLs (legacy)

note

Datasets can contain at most 1600 columns. For more information on how .csv files are processed during upload, refer to DuckDB docs.

Format

Load has several formats:

Datasets

Load data from the datasets <datasets>. Loads multiple datasets into your current session at once.

Files

warning

Load data from the file is a legacy skill and is no longer supported. Refer to Datasets for .csv and .xlsx uploads.

Load data from the file <file> loads data into your current session from a file that has been uploaded to DataChat. See Upload Files for more information.

Load data from the file <file> also supports the following optional qualifiers:

• using the delimiter <delimiter>

  From an uploaded file, specify a delimiter on which to split values. You can specify any string value, or can select from the following common values: comma ,, pipe |, semicolon ;, space, or tab.

• with columns <list of columns> having types <list of types>

  From an uploaded file, specify the types of specific columns. Note that this was most useful when DataChat didn't infer the column type you expected or preferred.

For .xslx files that include sheets, you can use two optional qualifiers to further specify which data to import. Use the qualifiers separately or together:

• excluding the sheets <sheet name>

  From an uploaded .xlsx file that includes multiple named sheets, exclude specific sheets from your current session.

• including the sheets <sheet name>

  From an uploaded .xlsx file that includes multiple named sheets, include specific sheets from your current session.

• with cell ranges <sheet name!cell range>

  From an uploaded .xlsx file that includes named sheets, specify cell ranges per sheet. Options for <cell range> include:

  • cell name: which loads all values from cell-name through the rest of the sheet
  • cell name:cell name which loads all values from cell-name to cell-name
  • cell name:column name which loads all values from cell-name up to and including all values in column-name

Snapshots

warning

Snapshots are a legacy object and are no longer supported.

Load data from the snapshot <snapshot name>. Loads the datasets from a snapshot into your session. A snapshot is a database that contains the datasets from another session that is created with the Save skill (either by you or by another user in your organization and shared with you).

URLs

warning

Loading from a URL is considered legacy and is no longer maintained.

Load data from the URL <URL>. Loads data directly from a URL endpoint and can split values using an optional delimiter. See the Files section for optional qualifiers.

Parameters

When loading data into DataChat, you can use the following parameters to specify where to load data from and how to load it:

• cell name (required) For an .xlsx file, the name of the cell name to include within a cell range.
• cell range (required) For an .xlsx file, the name of the cell range to include.
• column name (required) For an .xlsx file, the name of the column to include within a cell range.
• datasets (required) A comma-separated list of datasets to load.
• delimiter (optional). If specified, this parameter splits the values in the dataset on the specified delimiter. The options include a space, a tab, a comma, a semicolon, or a custom string.
• file (required). The name of the file to load.
• list of columns (optional). The columns to give specific types and formats when loading.
• list of types (optional). The types to assign to each of the columns from the list of columns parameter.
• recipe name (required). The name of the recipe to load data from.
• sheet name (required). For an .xlsx file, the name of the sheet to include or exclude.
• snapshot name (required). The name of the snapshot to load data from.
• URL (required). The URL from which to download the dataset.

Output

If the data is loaded successfully, the system returns a success message and shows a sample of the dataset.

If the data is not loaded, the system returns an error message in the conversation history.

Examples

To load data from the telcoCustomerChurn dataset, enter

Load data from the datasets telcoCustomerChurn

To load data from both the telcoCustomerChurn and Credit_Risk dataset, enter

Load data from the datasets telcoCustomerChurn, Credit_Risk