Defining correlation statistics

These steps are part of the larger task of Determining the linear relationship between variables in two columns. When you complete these steps, click the link at the bottom of the panel to return to the main task.

To define correlation statistics:

1. Right-click anywhere on the Transformer definition table and select Add. A row is added to the table.

2. Under the Data Column 1 heading in the row you just added, click and select the first column for the calculation. Only columns of numeric type are listed.

3. Under the Data Column 2 heading, click and select the second column for the calculation. Only columns of numeric type are listed. You cannot use the same column in Data Column 2 that you use in Data Column 1.

4. Double-click under the Statistics heading in the row you just created. The button is displayed.

5. Click the button. The Correlation - Select Statistics window opens.

6. Click one or more statistics from the Available statistics list. Then, click >. The statistics are moved to the Selected statistics list.

7. Repeat steps 1 through 6, as appropriate.

8. Click OK. The Correlation - Select Statistics window closes.

The Correlation transformer supports partial data. For example, if you select a column to define statistics for but don't select the statistics for it, the Correlation transformer will save your column selection. However, you cannot map columns for a row that has a partial data selection, nor can you successfully run a step that has a partial data selection.

Return to Determining the linear relationship between variables in two columns.