Tableau dataset builder

A package for programmatically creating Tableau data source (.tds) and packaged data source (.tdsx) files.

Method 1: Using a JSON manifest

The package turns a JSON structure representing a dataset into a .tds or .tdsx output.

The JSON model consists of:

• The link to the data source file (CSV or Excel)
• A link to additional metadata (optional)
• A list of measures
• A list of dimensions
• groups of fields to represent as folders
• hierarchies of related dimensions

There is an example JSON file in the example folder.

Below is a minimal example, using the sample data file and JSON model:

fromtableau_metadata.datasetimportcreate_tdsxcreate_tdsx(
metadata_file='example'+os.sep+'metadata.json',
dataset_file='example'+os.sep+'dataset.json',
data_file='example'+os.sep+'orders.csv',
output_file='output'+os.sep+'test_create_tdsx'
)

This will create a packaged data source from the example.

Method 2: Using the API to build a data source

You can also build data sources directly using the Tableau class.

The following is a minimal example of creating a .tds:

tableau=Tableau()
tableau.set_csv_location('example/orders.csv')
tableau.add_measure('Sales')
tableau.add_dimension('Ship Mode')
tableau.hide_other_fields()
tableau.save("output"+os.sep+"test_tableau.tds")

Method 3: Convert a CSV directly to a .tdsx

The simplest possible approach is to directly convert a CSV file into a Tableau Packaged Data Source in one step. This will include all the fields in the CSV without any additional metadata.

create_tdsx_from_csv(data_file='orders.csv', output_file='datasource')