Re-run a data mining run

Once a data mining run has been submitted, it can be re-run as is or with modified parameters. For example, you need an MGPS run with 5 as the Minimum Count, and another MGPS run with all of the same data mining parameters but with 10 as the Minimum Count. Once you have submitted the first run, instead of creating a new run and supplying all of the same selections again, except for the minimum count, you can re-run and change one parameter.

You can re-run any of the data mining runs that appear on the Data Mining Runs home page, including runs created by other users and runs that are currently running, completed successfully, failed, were canceled, or have not started. You can also re-run runs that are themselves re-runs.

Note:

• You cannot change the run type when re-running a run. For example, you cannot change from an MGPS run to a logistic regression run.
• To learn more about the error that caused a run to fail, review the Status column on the Jobs for Run page.
• The re-run option is not available for logistic regression runs created prior to the installation of Oracle Empirica Signal version 7.1, which included significant enhancements to logistic regression computations.

When you re-run a data mining run, Oracle Empirica Signal treats it as an independent run that is not associated with the original run; you are the owner of the new run. The run uses all parameters and options used for the original run (including the As Of date of the run, if the data configuration supports timestamped data) unless you explicitly modify them. In addition, the re-run uses the user preference, Replace Missing Values with, setting that was in effect for the run creator when the original run was created.

1. In the left navigation pane, hover on the Data Analysis icon (), then click Data Mining Runs.
2. From the Project drop-down list, select a project, or select -- to include all projects.
3. From the Configuration drop-down list, select a data configuration, or select -- to include all configurations.
4. Select the Row Action menu icon () for a run, and then click Re-run.
5. To select a different data configuration, from the Configuration drop-down list of compatible data configurations, select a data configuration.
6. To limit the run to only cases meeting specified criteria, check Add Database Restriction.
7. Click Next.

   If you added or modified a database restriction, the Define Database Restriction page appears. Otherwise, the Run Options page appears.

8. Modify run options as needed. You can also click Back to review all of the selections made for the original run and modify them as needed.

   Note:

   The run uses all of the variables, values, and parameters selected for the original run unless you modify them in the new run. For example, if there was a custom term for the original run, it is used as is unless you modify it. Keep in mind that any changes made to queries on the Queries page have no effect on database restrictions or custom terms that were part of the original run. During the re-run, you must modify the database restriction or custom term, if needed, including reselecting a query that was modified on the Queries page.
9. Click Next.
10. On the Name Data Mining Run page, in the Run name field, enter a name for the run. The name of the run does not need to be unique, although we recommend that you use a unique name. Oracle Empirica Signal assigns a unique Run ID to each run.
11. In the Description field, optionally enter a description. Oracle recommends that you provide an informative description of the run so that users who view run results know which run to use.
12. (Optional) To assign the run to a project, choose one of the following:
    • To assign the run to an existing project, click Add to existing project and select from a list of projects associated with objects that you created or that are published to you.
    • To create a new project and assign the run to it, click Add to a new project named and enter a project name.
13. Click Next.
14. On the Confirm Run Parameters page, review the parameters chosen for your run to make sure that they are correct. Different parameters display for MGPS runs and LR runs.
15. If you want to change any parameters, click Back until you are on the appropriate page and make the necessary changes. Then click Next until you are back on the Confirm Run Parameters page.
16. When you are satisfied with the run parameters, click Submit.
17. Click Continue.

The run status appears on the Data Mining Runs page. To monitor the progress of a run, you can view jobs for the run.