Oracle® Retail Size Profile Optimization User Guide for the RPAS Fusion Client Release 14.1 E56467-01 |
|
![]() Previous |
![]() Next |
The Location Clustering workbook allows users to cluster locations based on similarity of their size profiles to create ideal prepacks for each of the clusters.
Size clusters in the context of SPO are useful for determining prepacks that are best suited to each cluster and may be used in downstream applications such as Assortment Planning and Allocation.
The Location Clustering task contains the Location Clustering step.
Creation of location clusters involves the following steps:
Use the Location Clustering task within SPO to create location clusters for each size range.
Create up to ten clusters based on location size profiles for each size range and attribute.
The clusters for each size range and attribute combination are independent of one another.
Use the list in the Number of Clusters view to select the number of clusters required for the class.
The Location Clustering option creates location clusters and assigns each of the selected locations to one of the clusters.
To build a Location Clustering workbook, perform the following steps:
Click the Create New Workbook icon on the Location Clustering task.
You may be prompted to select a domain if you have access to more than one domain. For information on domain selection, see "Selecting the Domain when Creating a New Workbook" in Chapter 2.
The Select Class workbook wizard is displayed. Select the class for which the location clusters have to be created, and click Next.
The workbook wizard opens at the Select Store screen. Select the locations and click Next.
The Location Clustering workbook, with the Location Clustering step, is built.
The Location Clustering step has the following views:
This view is used to define the number of clusters required for the selected class.
The following table describes the measures available on this view.
Table 6-1 Numbers of Clusters View Measures
Measure | Description |
---|---|
Number of Clusters |
Select the number of clusters required for the class. |
After you have specified the number of clusters, use the Location Clustering planning action to perform the following:
Create as many location clusters as needed based on the centroids of the size profiles.
Assign each of the locations to one of the available clusters, based on the relationship of the location's size profile to that of the cluster.
This view displays the escalated profile that was eventually approved. This is the size profiles that would be used in Location Clustering in the next view.
The following table describes the measures available on this view.
This view displays the centroids of each of the clusters. A centroid is the average of the members in that cluster. This view is used to view the profiles of the clusters that would be created.
This view shows the assignment of each of the locations to one of the clusters for the selected class.
The example here shows that a given location can belong to different size clusters for the same class of merchandise-based on the size range and attribute.
The following table describes the measures available on this view.
This view displays the total number of locations in each of the clusters for every size range and attribute within the class.