Splitting a dataset into a training and test set

In this recipe, you will split the data into training and test sets using the SSIS percentage sampling transformation. You will use 70 percent of the data for the training set and 3 percent for the test set.

Getting ready

There are no special prerequisites for this recipe, except, of course, SSIS 2016 installed, and the AdventureWorksDW2014 database available in your SQL Server instance.

How to do it...

Open SQL Server Data Tools (SSDT) and create a new project using the integration services project template. Place the solution in the C:\SSIS2016Cookbook folder and name the project Chapter08:

1. Rename the default package to SplitData.dtsx.
2. In the Control Flow tab in the Package Designer, add a new data flow task by dragging and dropping it from the SSIS toolbox.
3. Right-click the task and select Rename from the pop-up menu. Change the task's name to SplitData.
4. Click the Data Flow tab.
5. Create a new OLE DB source. Name it AW_DW_Source.
6. Double-click the AW_DW_Source...