Customizing Each Segment

Working with the Profile Node

Right-click the profile node for your pipeline and select Open.

The profile view includes the Advanced Filter pane and a table listing each segment.

Advanced Filter

You can filtera set of specified criteria that are applied to data in order to identify the subset of data for a subsequent operation, such as continued processing. the displayed segments by segment name or by the number of time seriesan aggregation of transactional data into specified time intervals and sorted according to unique combinations of the default attributes (BY variables) in each segment.

Segment table

The table listing the segments shows the segment name and the number of time series within each segment. For External Segmentation, the segment names come from the _SEG_ value in the data source.

• In the Demand Classification Profile, there are 11 segments. The name is determined by patterns detected in the time series and classified by the _DEMAND_CLASS attribute. For more information, see Demand Classification.
• In the External Segmentation Profile, there can be any number of segments. The segments are listed based on the value of the _SEG_ attribute, which is where they are defined in the data source.

1. To change the name of a segment, right-click the segment and select Rename. The name that you provide is used for this pipeline only.
2. When you are finished making changes, click Close.

Overview of the Modeling Node

Right-click the modeling node for your pipeline and select Open. The segments are listed in the table on the right. Each segment in the pipeline has its own nested pipeline.

Advanced Filter

You can filter the segments shown based on the following columns:

• _DEMAND_CLASS or _SEG_
• Number of Time Series
• Status
• MAPE

Use these tools to customize the attributes displayed in the Advanced Filter.

Opens Filter Settings for selecting which attributes should be displayed in the Advanced Filter, and in what order.

Expands each attribute so that all values are visible.

Collapses each attribute so that no values are visible.

These attributes are explained below.

Segment table

The following icons are listed above the segment table. If you right-click a segment, each each of these actions are available in a context menu.

• ─ Runs all of the pipelines that you have selected.
• ─ Opens the Browse Templates window. You can replace a pipeline that you have selected with one of the pipeline templates from The Exchange
• ─ Opens a pipeline that you have selected. . This icon is enabled when only a single pipeline is selected.
• ─ Resets selected pipelines to use the default pipeline.

The table listing the segments shows following information for each segment.

Segment Name

By default, the name of the segment comes from the source. For External Segmentation, the source is the value of the _SEG_ variable from the imported attributes. For Demand Classification, the source is the value of the _DEMAND_CLASS variable, based on the characteristics of the time series.

The segment name can be changed in the profile node.

_SEG_ or _DEMAND_CLASS

The original value of the segment name. This name cannot be changed in the pipeline.

Number of Time Series

This column shows the distribution of time series for all of the segments.

It is common to have some empty segments in the Demand Classification pipeline.

Pipeline

This is the nested pipeline that is assigned to each segment.

• External Segmentation

  Each pipeline is assigned using the Default pipeline setting for the External Segmentation Modeling node. If the value is changed after running the pipeline, it does not change the assigned pipeline for each segment. To change the assigned pipeline to the new default setting, select one or more segments and click .

  After successfully running the segmented pipeline, if you reset the default pipeline, the modeling node and the following nodes show a status of Out-of-date. If you open the modeling node, the individual segments show a status of Completed. However, the nodes that are Out-of-date should be run again.

• Demand Classification

  Each pipeline is determined by the patterns detected in each time series, using the best template for each classification. This can be changed to any other template, including any custom templates available for your site. You can also open the pipeline to modify the settings to the modeling node, or add other modeling nodes.

To make specific changes to the pipeline for a segment, see Customizing Pipelines for Each Segment. If you make any changes to a segment’s pipeline, the pipeline name indicates Custom in the modeling node.

Note: The Copy pipeline link action is not available for nested pipelines. For more information, see Copy pipeline link.

Status

Each segment has one of the following status indications.

• Created — This value is assigned to all segments when a segmented pipeline is first created. The Status cell is empty.
• Pending — The pipeline is waiting for other processing to complete before it can start execution.
• Running — The pipeline for the segment is running.
• Modified — A change has been applied to the pipeline for the segment. The pipeline must be run again.
• Failed — The pipeline has completed with errors. Open the pipeline for the segment, check the location of the error, and open the log.
• Canceled — Pipeline execution has been canceled.
• Completed — The pipeline run has completed successfully.
  Note: After running one or more segments in the modeling node, segments with zero time series show a status of Completed if the other segments completed successfully. This is necessary for the modeling node to attain a status of Successful.

  Subsequent data refreshes can populate the segment with time series. You can open the pipeline for an empty segment and modify it. This changes the segment’s status to Modified. To return the segment’s status to Completed, right-click the segment and select Run pipeline.

  Always run segments with zero time series from the modeling node. If you open the segment’s pipeline and run it, the pipeline fails.

Customizing Pipelines for Each Segment

Any changes you make to the pipeline for a segment changes the segment’s status to Modified. The pipeline must be run again.

To make changes to the pipeline for each segment, follow these steps.

1. Right-click the External Segmentation Modeling or Demand Classification Modeling node for your pipeline and select Open.
2. To choose another pipeline template for the segment, right-click the segment and select Replace pipeline. The Browse template window is opened and you can select the replacement template for the segment.
   Note: You can replace multiple pipelines that you have selected by using .

   Some pipeline templates in The Exchange are not available for use in a segment pipeline. In External Segmentation Modeling, you can replace the pipeline for a segment with Demand Classification. The External Segmentation template is not available as a replacement for any segment. In Demand Classification modeling, segmented pipelines are not available for replacement.

3. To customize the pipeline for the segment, right-click the segment and select Open pipeline. The nested pipeline for the segment is opened. You can add and remove nodes to this pipeline.
   1. To add other modeling nodes to the pipeline, right-click the Data node, select Add child nodeForecast Modeling.
   2. To remove a modeling node, right-click the node and select Delete.
   3. To change how each modeling node uses eventsan incident that disrupts the normal flow of any process that generates the time series. Examples of events are holidays, retail promotions, and natural disasters., right-click the node and select Modify event usage.
   4. When you are finished making customizations, click Close. You are returned to the modeling view and the list of segments. The segment’s status changes to Modified.
4. To run a pipeline for a segment, right-click the segment and select Run pipeline.
5. After making changes to the pipeline for a segment in External Segmentation Modeling, you can change it back to the default pipeline by right-clicking the segment and selecting Reset to default.
   Note: You can reset multiple pipelines that you have selected to use the default by using .

   For Demand Classification, do not use Reset to default. This sets the pipeline to Auto-forecasting. The default assignments for each segment is listed in Demand Classification Templates.

6. When you are finished making changes to the segments, click Close to return to the segmentation pipeline.