Enabling frequency based bucketing
Frequency based bucketing counts the number of members in each bucket, and then determines the buckets that exceed the maximum bucket frequency number.
About this task
Use this procedure to enable frequency based bucketing on a particular bucketing group. You must be in the Configuration perspective.
- In the toolbar, select Advanced Interface from the Editor interface list.
- In the configuration editor, select the Algorithms view.
- Select the bucketing group.
- In the Properties view, set the Maximum bucket size property in the Properties view to a value greater than 0. If Maximum bucket size is greater than 0, Maximum attribute tokens must be greater than Minimum attribute tokens.
- Redeploy the algorithm to the InfoSphere® MDM instance.
- Generate your frequency files by running the Generate Frequency Stats (mpxfreq) utility. Set the Generate frequency tables for frequency-based bucketing option. The output of mpxfreq must be copied into the project from InfoSphere MDM through the Jobs view by running the Get job results action on the successful Generate Frequency Stats (mpxfreq) job.
- Deploy the configuration to the InfoSphere MDM instance.
Remember to regenerate derived data before continuing with weight generation.