Exploring bivariate value distributions

By using the bivariate statistics function, you can select one categorical column in a table as the target field and then explore the value distributions of the other fields for each distinct value of the target field.

The Intelligent Miner clustering visualizer is used to display the bivariate statistics results.

Each value of the target field, for example MARITAL_STATUS=Married, is displayed as a row/cluster in the clustering visualizer. The charts for a cluster show the value distribution of all other fields for records that have a value of MARITAL_STATUS=Married.

To compute bivariate statistics for a field, follow these steps:

In the Data Source Explorer, right-click a table, for example, BANKCUSTOMERS, and click Distribution and Statistics > Bivariate Distribution.
On the Select Target Column window, click a column, for example, GENDER, and click OK. The Clustering Visualizer is displayed and shows the bivariate statistics for the column GENDER.

When you start the calculation, a progress dialog is displayed. With this progress dialog, you can monitor the progress of the calculation, set the calculation in the background, or terminate the calculation by clicking Cancel. If you click Cancel, the calculation terminates after a reasonable time. After the calculation is completed, the result is displayed.

You can view the progress of the calculation or cancel the calculation also in the Execution Status view.

Feedback