Tip:
Highlight text to annotate it
X
Welcome to "Creating charts in Oracle Endeca Information Discovery Studio using data from multiple views".
If you are new to Studio, then you should probably start with the "Getting Started with OEID 3.1" screencast series.
Before viewing this video, you should already be familiar with:
How to create applications and data sets
How to add components to a page
The basic configuration for a regular Studio Chart component
In this video, I'll briefly discuss how refinement rules allow you to link attributes from different data sets.
Studio uses data sets to create base views.
I'll then show you how to configure a chart to display data from more than one view,
and show how Studio uses refinement rules to help select the chart dimensions.
This video is intended for users who create applications using multiple related data sets,
and who want to be able to show data from those data sets on a single chart.
Why use multi-view charts?
Multi-view charts allow you to directly compare metric values from different data sets,
such as in this example, where a chart shows both sales numbers and tweet numbers.
Otherwise, bringing these metrics together would require you to create additional custom views.
To create a multi-view chart, the views must have dimensions with the same data type and values.
Refinement rules allow a refinement from one data set to be replicated for an attribute in a different data set.
For example, when you refine the sales date by a specific date range, the tweet date is also refined by that date range.
Multi-view charts can use refinement rules to help determine whether two views should be used for the same chart.
For attributes to be linked by a refinement rule, they must:
Have the same data type
Be all single-value, or all multi-value
Have the same refinement behavior
When you add a data set to an application, Studio can automatically create refinement rules between attributes that
have the same name,
and that meet the other refinement rule requirements.
For this video, we've created an application that uses data from a fictitious bicycle manufacturer.
The data sets for the application include:
Sales records
Product information
Employee information
Reseller information
The Sales records reflect information from the other data sets. The data sets are linked using refinement rules.
On the Refinement Rules page for the application, you can see that the data includes refinement rules for several attributes,
including for product category and product subcategory.
Let's go to the application to see one of these refinement rules in action.
On this page, the Sales page, the Available Refinements component shows the available refinements for sales.
We also have two Selected Refinements components – one for sales, and one for products.
Let's refine the sales data by the Road Bikes product subcategory.
Because of the refinement rule, you can see that Studio has automatically added refinements to the Sales data and the Products data.
When you remove the refinement from one data set, the linked refinement also is removed.
When we configure our multi-view chart, you will see how this refinement rule helps with the configuration.
Let's configure the default chart on the Sales page to display both product and sales data.
On the chart edit view, to indicate that you want to use multiple views, check the Enable charting across multiple views checkbox.
The radio buttons for each view change to checkboxes.
The Sales view is already selected.
To add the Products view to the chart, I'll check the Products checkbox.
When you are using multiple views, Studio displays the list of selected views at the top of the edit view.
On the other tabs, to remove a selected view from the chart, click its delete icon.
You can use multiple views for any type of chart.
I'll leave the default chart type, a basic bar chart.
On the Chart Configuration tab, the Attributes list contains the available attributes from all of the selected views.
Each attribute name is followed by the name of the view it belongs to.
You can filter the list to only include attributes from a specific view.
For the chart dimension lists, each dimension is actually a dimension group.
For each dimension group, you must select a dimension from each view.
Let's start by adding our metrics. I'll filter the list to only show the Sales attributes.
I'm selecting the Unit Price metric from sales, and setting the aggregation to average.
Now I'll filter the list to only show product attributes.
I'm selecting the List Price metric, and adding it to the same metric group as the Unit Price metric.
When you add both metrics to the same metric group, they're displayed at the same time on the chart.
Because these are both currency values, I'll configure the metric group so that the values for both metrics are displayed as currencies.
You can also display attributes from different views as separate metrics, and allow end users to switch between them.
I'll also add the same List Price and Unit Price metrics as separate available metrics for the chart.
For the dimension, I want to use the Product Subcategory dimension.
I'll start by dragging the Sales_ProductSubcategoryName dimension from the Sales view.
When I expand the dimension group, you can see that Studio has automatically added the Product Subcategory Name
dimension from the Products view.
This is because these attributes are linked by a refinement rule.
If a refinement rule isn't defined, you can also manually add the dimensions to a dimension group.
For dimension groups, you can customize the display name.
By default, the chart uses the name of the first dimension that you added.
I'm going to update the display name to be Product Subcategory, which works for both views.
Let's save the chart, and exit the edit view.
On the end user view, the default chart metric is the combination of both the Unit Price metric, from sales,
and the List Price metric, from products.
You can see on the chart that the List Price is generally higher than the Unit Price, and often by about the same percentage.
You can also see where a product subcategory is missing one of the values.
I can also switch to one of the individual metrics that I added.
When I change to a metric from a different view, Studio switches seamlessly,
changing automatically to the correct dimension for that view.
Now I'll refine by the Components product category in the sales data.
Because of the refinement rule for that attribute, Studio adds refinements for both the sales and product data.
Note that if your dimensions are not connected by a refinement rule, then only the one view is affected.
The chart is also updated accordingly, with the correct subcategories shown for each metric.
So to recap:
Refinement rules allow you to link refinements to attributes from different data sets.
Refinement rules can help with the configuration of multi-view charts.
On a multi-view chart, you can display metrics from different views at the same time,
or allow users to switch between individual metrics.
Each dimension is actually a dimension group, with a dimension from each view.
For more information on refinement rules and multi-view charts, see the Studio User's Guide.