Oracle^® Business Intelligence Applications Installation and Configuration Guide > Configuring Oracle's Supply Chain Analytics family of products > Process of Configuring Supply Chain Analytics for a Universal Source >

About Configuring the Purchase Cycle Lines Aggregate Table

To aggregate the Purchase Cycle Lines table (W_PURCH_CYCLE_LINE_A), you need to configure the parameterfiledw.txt parameters file and the source system parameters file, and run the initial ETL workflow and then the incremental ETL workflow.

For your initial ETL run, you need to configure the GRAIN parameter for the time aggregation level in the Purchase Cycle Lines Aggregate fact table.

For the incremental ETL run, you need to configure the time aggregation level and the source identification. The source identification value represents the source system you are sourcing data from.

You need to configure two parameters to aggregate the Purchase Cycle Lines table for your incremental run:

• GRAIN
• $$TIME_GRAIN

These parameters has a preconfigured value of Month. The possible values for parameters are:

• DAY
• WEEK
• MONTH
• QUARTER
• YEAR

The Purchase Cycle Lines aggregate table is fully loaded from the base table in the initial ETL run. The table can grow to millions of records. The Purchase Cycle Lines aggregate table is not fully reloaded from the base table after an ETL run. Oracle Business Analytics Warehouse minimize the incremental aggregation effort, by modifying the aggregate table incrementally as the base table gets updated. This process is done in four steps:

1. Oracle Business Analytics Warehouse finds the records to be deleted in the base table since the last ETL run, and loads them into the W_PURCH_CYCLE_LINE_TMP table. The measures in these records are multiplied by (-1). The mapping responsible for this task is suffixed with _Derive_PreSoftDeleteImage, and it is run before the records are deleted from the base table. It is run in the source-specific workflow.
2. Oracle Business Analytics Warehouse finds the records to be updated in the base table since the last ETL run, and loads them into the W_PURCH_CYCLE_LINE_TMP table. The measures in these records are multiplied by (-1). The mapping responsible for this task is suffixed with _Derive_PreLoadImage, and it is run before the records are updated in the base table. It is run in the source-specific workflow.
3. Oracle Business Analytics Warehouse finds the inserted or updated records in the base table since the last ETL run, and loads them into the W_PURCH_CYCLE_LINE_TMP table, without changing their sign. The mapping responsible for this task is suffixed with _Derive_PostLoadImage, and it is run after the records are updated or inserted into the base table. It is run in the post load-processing workflow.
4. Oracle Business Analytics Warehouse aggregates the W_PURCH_CYCLE_LINE_TMP table, and joins it with the W_PURCH_CYCLE_LINE_A aggregate table to insert new or update existing buckets to the aggregate table. This step is part of the post load-processing workflow, and the mapping is suffixed with _Derive.