This document discusses several topics which are good to understand when working with the OB Analytics window.
OB automatically handles currency conversions automatically when creating and updating fact tables.
The implemented solution automatically duplicates each fact for each currency. This means that one sales fact is present for each currency, using the relevant date (order date for example) for the conversion rate.
As each real fact is translated in one or more analytics fact it means that it is critical to always include the currency dimension within the MDX/analytics setup as a dimension or filter of the MDX query. If you don't include the currency dimension you will see duplicated sales/quantity values added together across different currencies.
Several types in Openbravo can be structured as a tree. For example product category or business partner category. Openbravo out of the box supports trees in the Mondrian integration. However, because of performance reasons this is not used/setup for the standard dimensions in the cubes.
This is done by checking a flag in the dimension definition. If you for your case need the dimension with tree support it is quite easy to do: add a new dimension mapping to the same property and with the above flag not checked.
For the same reason the product and product category hierarchy and the business partner and business partner category hierarchy has been mapped as separate dimensions. This has consequences for the way they are visualized in the UI, as separate dimensions instead of one combined dimensions.
However, for convenience for both hierarchies there is also a combined dimension:
- BP Category Hierarchy
- ProductCategory Hierarchy
which allows you to use one dimension in the grid/chart combining both the bp and the bp category structure and the product and product category.
Date Time Mapping
Openbravo provides two hierarchies for each date dimension. The first hierarchy allows you to zoom in from year to day level step-by-step. The second hierarchy allows you to directly include a date column/row in the visualization.