Using spatial reporting in OBIEE is not for the faint of heart. However it is also rewarding. The reports look great; it is new and henceforth exciting. It is challenging but the outcome stands out. However on the route to the reports, some obstacles must be overcome. Let me provide some thoughts on the obstacles that I encountered.
The first obstacle is the decision what tools to use. One must use next set of tools: [a] the mapbuilder which is an independent java tool that can be run from any client and which creates some layers that will be shown on a map, [b] the mapviewer which is used to create a map with the layers that were created by the user and [c] the “manage map data” that can be found under the “map data management” in the administration of Oracle BI presentation. To know when to use what tool was not trivial. First, the layers must be created in the mapbuilder; then one must decide on the map (to be set up in the mapviewer) and finally the map must be linked to the data in the manage map data part. It took me some time to figure this out.
Then the decision on how to set up the maps. Do we take predefined map (such as the Bing map, the Google map, the Navteq map) or do we create our own maps? Let me provide an example on a map that is created on top of an existing predefined map:
and then an example with a map that is created by yourself:
.
This choice has some consequences. Do you accept the default settings that are more or less imposed by the predefined map or do you want to set your own standards, labels, colours etc? Both choices have pros and cons. Something to think about. If one settles for a own map creation, one might consider using this manual. It helped me to create some maps that I created myself. If you use the predefined maps, it is somewhat trial and error. But it worked out for me.
Another choice is the model to use. One must create a model that is propagated from physical layer to the presentation layer and that can be found in the subject area. In the subject area, we must have some labels that refer to either points in space or areas in space. As an example, one must have labels that refer to shops and which can be seen as points and municipalities that are areas and for which we must have labels. Such labels are important. Somewhere down the line we must link these labels (shops, municipalities) to spatial objects (points, polygons) in order to render them on screen.
A final choice is where the spatial data (longitudes/ latitudes) might be derived from. Do we download them/ Do we copy them from an existing source or do we create them ourselves. Everything is possible. In short: choices, choices, choices!
