Combining Data Sources in one Visualization

    Sometimes you need Data Blending to extract value from more than one data source in the same Analytics visualization.

    For example, you may want to compare vacation days taken with other employee absences by department, to gain insight about the correlation of those variables.

    Data Blending Example Visualization

    Follow these steps to combine two data sources in one visualization:

    1. Open the Data Blending dialog.

      Click/Tap the + button in the Fields section and select Fields from another Data Source.

      Data Blending Accessing dialog in the Visualization editor

    2. Choose the new Data Source.

      Connect to the data source that you want to combine.

    3. Specify the JOIN Condition

      Configure the equality condition that needs to match to combine the two data sets.

      Data Blending Join Condition selection
    4. Choose the fields you want

      Specify which are the fields that you want to combine, so you can access them in your visualization.

      Data Blending Choose New Fields option
    5. Select Join Data

      After merging the two data sets, you can find the new fields at the bottom of the Fields section.

      Show Data Blending Result

      As shown in the image above, you can now visualize vacation days by Department name instead of using the internal Department ID.

    The JOIN Condition

    When adding fields from other data sources, you actually join two different data sets. The join operation used by Analytics is LEFT (OUTER) JOIN.

    Below you can see how the Vacation (left table) and the Departments (right table) data sets are joined, using the DepartmentId field in both tables as the relationship between them (equality condition: DepartmentId = DepartmentId).

    Before JOIN operation:

    Vacation

    Departments

    Taken (days) DepartmentId

    40

    1

    92

    10

    DepartmentId Department (name)

    1

    CPA

    100

    HR

    After JOIN operation:

    Taken (days) DepartmentId DepartmentId Department (name)
    40 1 1 CPA
    92 10

    Notice that LEFT JOIN operation returns all records from the left table, and keeps only matching records from the right table.