About Graphs and the Business Services Data Model in JClient

The JClient oracle.jbo.uicli.jui package includes the following classes which are related to the databound Graph:

JUGraphBinding
JUSingeTableGraphBinding
JUMasterDetailGraphBinding

To understand the selections you make when building the graph component, you must understand how the Graph bean obtains sufficient data from your business components data sources to:

Plot each graph marker (as determined by up to three data points)
Label each graph group; for example, the months of the year
Draw either a single series or a multi-series type graph
Label each graph

The data binding is determined in the Graph Customizer by the data model components (e.g. Business Components view object attributes) you select. The Graph Customizer helps you choose objects and attributes for a particular graph type, but you must be familiar with the data model in order to make the specific choices. In general, when you make a graph type selection in the Graph Editor, the Graph Customizer displays one attribute for each data point value needed to draw the markers for your graph type. If you select a group type graph, the Graph Customizer also prompts you to choose an accessor that links the master data collection with the correct detail data collection.

For instance, if you want to plot a Hi-Lo-Close Stock graph type for multiple stocks in the Graph Customizer, you must choose:

The master data collection that contains one row for stock type
The detail data collection that contains the rows corresponding to the data of each stock
The attribute from the master data collection that specifies the label for each stock graph
One attribute from the detail data collection for each marker data value (in this case, Hi, Low, and Close)

Whereas, if you decide to plot a single stock as a simple bar graph that shows monthly high values in the Graph Customizer, you must choose:

An unconstrained data collection (one that does not belong to a master-detail relationship)
One attribute that provides the label for each month
One attribute for the marker data value (in this case, Hi value only)

The following section describes the business components data model requirements in detail.

Business Components Data Model for Graphs

It is helpful to understand how the Graph bean interprets the data from the business services data model. In general, the requirements of the data model for your ADF Model project depend on:

Whether the graph type you select requires one, two, or three data point values to determine each graph marker
Whether you want to plot a series (single table) or a group type (master-detail) graph

For example, let's assume a simple graph, such as a Bar graph with the following data is desired:

 

ENAME   SAL     Comm 

KING     1000    200 

CLARK    2000    100 

MILLER   1500    50

Each row in this table will correspond to a Series in the graph and each column corresponds to a Group.

Some types of graphs, like the Bar graph, require one value per marker. This is in contrast to other graph types, like the Stock HLC graph, which require three values per marker (high, low and close). When your graph requires multiple data values, it is convenient to store them in separate rows in your database table:

 

Date         High  Low Close 

10 Jun 03    11    10   11 

11 Jun 03    11     7    9 

12 Jun 03    10     9  9.5

The Graph beans supports the type of graph whose data is stored in a single table. When your graph requires multiple values, then the data for the chart should be modeled as a master/detail relationship. Each detail provides one series of data, corresponding to the master value. In the Stock graph example above, the data model could look like this:

  

Master table stock_ticker_table 

-------------------------------- 

ticker             symbol 

Oracle Corporation ORCL 

XYZ Corporation    XYZ 



Detail table stock_price_table 

----------------------------------- 

ticker  Date       High  Low Close 

ORCL    10 Jun 03    13   12   13 

ORCL    11 Jun 03    14   13   13 

ORCL    12 Jun 03    15   14   14 

XYZ     10 Jun 03    10    9    9 

XYZ     11 Jun 03    10    9    9 

XYZ     12 Jun 03    10    8    9

In summary, the case of the single-table graph, the data model is simple and need only contain:

A standalone data collection, where each row plots one graph marker
One attribute that supplies the label for the groups that the graph displays
One attribute per graph data point, with as many attributes as required to plot each graph marker

In the case of the more complex group type graph, the data model is based on a master-detail relationship and must contain:

A master data collection, where each row specifies one series but contains no data
One master data collection attribute that supplies the label for each series
One detail data collection for each series, where each row plots one graph marker for that particular series
One detail data collection attribute that supplies the label for the groups displayed by the graph
One detail data collection attribute per graph data point, with as many attributes as required to draw each graph marker