Getting Started with Tableau 2018.x
上QQ阅读APP看书,第一时间看更新

Normalized extract (2018.3)

Previously, when creating an extract, Tableau always generated a single table. This was sometimes problematic, especially when a Join duplicated the number of lines. Now, when you Join multiple tables, you can choose the schema, creating either a Single Table or Multiple Tables

For example, you can download the Excel file Multiple Table Storage Test from the Chapter 1 section of my website, book.ladataviz.com (or, browse to https://ladataviz.com/wp-content/uploads/2018/09/Multiple-Table-Storage-Test.xlsx). This file contains the an Orders sheet, with approximately 10,000 rows and another sheet, User Access, with 6,000 rows.

Individually, those tables are small, and creating an extract only takes a few seconds. When you Join the two tables, the data is duplicated. The result of the Join produces more than 11 million rows. When you created an extract, prior to Tableau 2018.3, the 11 million rows had to be retrieved, as you can see in the following screenshot:

Thanks to Tableau 2018.3, if you select the Multiple Tables schema when creating an extract, the extraction time is very short because the two tables are generated separately, prior to being joined. The only drawback is that you can't use all of the Extract options (Filters, Aggregation, and Number of rows).

The next new feature is also related to Joins.