Defining relationships between attributes
Attribute relationships are associations between attributes that specify
how attributes are connected. Attribute relationships define how tables
and columns are joined and used, and which tables are related to other
tables. Without relationships, there is no interaction between data, and
therefore no logical structure. The relationships give meaning to the
data by providing logical associations of attributes based on business
rules.
Every attribute relationship has two parts—a parent and a child. A child
must always have a parent and a parent can have multiple children. The
parent attribute is at a higher logical level than the child is. For example,
in a relationship between Year and Quarter, Year is the parent attribute
and Quarter is the child.
Relationship types are defined by the attribute elements that exist
in the related attributes. Each type is described below:
One-to-one:
Each element in the parent attribute corresponds to one and only one
element in the child attribute, and each child attribute corresponds
to one and only one element in the parent attribute. A common example
of a one-to-one relationship is citizen and taxpayer ID. A citizen
can have only one taxpayer ID and a taxpayer ID can be assigned to
only one citizen.
One-to-many:
Each element in the parent attribute corresponds to one or more elements
in the child attribute, and each child attribute corresponds to one
and only one element in the parent attribute. These are the most common
types of attribute relationships. Year has a one-to-many relationship
to quarter. One year has many quarters, but a specific quarter can
be in one year only. This assumes that quarters are defined with an
accompanying year such as Q4 2006, Q1 2007, and so on.
Many-to-one:
Each element in the parent attribute corresponds to one and only one
element in the child attribute, and each child attribute corresponds
to one or more elements in the parent attribute. Many-to-one relationships
are the same type of relationship as a one-to-many, but it is defined
from a different perspective. For example, year is described above
as having a one-to-many relationship to quarter. This means that quarter
has a many-to-one relationship to year.
Many-to-many:
Each element in the parent attribute can have multiple children and
each child element in the child attribute can have multiple parents.
In banking, customers and accounts are an example of a many-to-many
relationship. One customer can have many accounts, and each account
can be associated with many customers, such as in the case of a joint
checking account.
With each attribute, the relationship type can be set to auto-managed
or user-managed. When a new dataset is imported, the application detects
the existing relationships between attributes and marks these relationships
as auto-managed. Any new data a user adds after the initial data import
is marked as user-managed. At any point in time, the type of relationship
management can be switched from auto-managed to user-managed, and vice
versa.
Prerequisite
The following steps assume that you have already
begun importing data. The Preview page is displayed after you select a
data source to import from, select the data to import, and click Prepare Data. For a list of the
data sources that you can import from, and links to steps to import data,
see About importing data.
For steps, see below:
To
define relationships between attributes
To
set relationships as user-managed or auto-managed
To define relationships
between attributes
On the Preview page, hover the
cursor over the name of a table, click the Menu
icon, and select Define Relationships.
The Define Relationships dialog box opens.
To add a new relationship, complete
the following steps:
Click
Add a New Relation. A blank
relationship is displayed.
From
the drop-down list in the Parent
Attribute column, select the parent attribute. The parent attribute
is at a higher logical level than the child is
From
the drop-down list in the Child Attribute
column, select the child attribute.
By default,
the relationship is defined as one to many. You can select a different
relationship type from the drop-down list in the Relationship
column:
1:1:
Each element in the parent attribute corresponds to one and only one
element in the child attribute, and each child attribute corresponds
to one and only one element in the parent attribute.
1:Many:
Each element in the parent attribute corresponds to one or more elements
in the child attribute, and each child attribute corresponds to one
and only one element in the parent attribute.
Many:1:
Each element in the parent attribute corresponds to one and only one
element in the child attribute, and each child attribute corresponds
to one or more elements in the parent attribute.
Many:1:
Each element in the parent attribute corresponds to one and only one
element in the child attribute, and each child attribute corresponds
to one or more elements in the parent attribute.
You can edit existing relationships,
by changing the Parent Attribute,
Child Attribute, and Relationship options.
To delete all the relationships
for the table, click Delete All.
When you have defined all the
relationships for the table, click Save
to return to the Preview
page.
To set relationships
as user-managed or auto-managed:
On the Preview page, hover the
cursor over the name of a table, click the Menu
icon, and select Define Relationships.
The Define Relationships dialog box opens.
There are two ways to manage attribute
relationships:
Auto-managed:
This is the default relationship management type when importing new
data. The application manages the relationships according to how elements
and attributes relationship types are set. If an attribute does not
match the relationship criteria that is set, the attribute is automatically
deleted. This eliminates duplicate attributes in your dataset.
User-managed:
This relationship management type allows a user to oversee and manage
the relationships between attributes and elements. This is recommended
for advanced analysts who are comfortable with Project attributes,
as the application does not automatically manage the imported datasets.
To select a management type for
each attribute, click the icon to the left of the Parent Attribute
To select a management type for
all attributes, click Convert All and select an option
from the drop-down list.
If user-managed is selected and the attribute
relationships are invalid, an error message is displayed and you won't
be able to save the relationships between the highlighted attributes and
elements.
Click Save to
make your changes.
In the future, you can edit the management types for attributes relationships
by click Edit Dataset from the drop-down menu under the
name of the dashboard dataset.
Related topics
Previewing
your data and specifying data import options
About importing
data
_____________________________
Copyright © 2019, Oracle and/or its affiliates. All rights reserved.
Legal Notices