You can find (discover) potential attribute joins between multiple entities. Each potential join shows how an attribute from one entity relates to an attribute from another. You can then save those joins that are important to your analysis and discard the ones that you do not need.
To find potential joins
-
Do one of the following:
- From the Home tab, click Joins > Discover.
- In the Navigation View, from the Analysis tab, right-click Joins and select Discover Join or right-click a Baseline Analysis project and select Analyze > Discover Join.
The Discover Joins window opens.
- In the Findings Name field, enter a name for your join results.
-
Expand each entity and check the box(es) next to the attributes to include in the join.
Note: Placing a check beside an entity name selects all the attributes in that entity.
- (Optional) Click Clear to remove the all entity/attribute selections and start again.
-
Configure the options using the information in the following table:
Option
Description
Match Quality %
Trillium only finds potential joins greater than or equal to the value entered for Match Quality %.
This value corresponds to the List View column Match Best. For example, to find all joins that match at least 90% of the time, enter 90% at the Match Quality %.
Include inferred M:M joins
Select to discover all potential joins (including many to many relationships).
If all attributes are selected for analysis, it is a recommended to clear this option to ensure best performance.
Self Joins
Select to join on attributes in the same entity.
Example: If an entity contains the following columns of information: EMP_ID, NAME, MANAGER_ID, you may need to join within that entity on EMP_ID and MANAGER_ID to determine which employees report to that manager.
Function
Trillium is not limited to joins solely based on data values. You can select from the following function types:
value
Attempt to join on attribute values
metaphone
Attempt to join on attribute metaphone values.
pattern
Attempt to join on attribute pattern values.
mask
Attempt to join on attribute mask values.
standardize - Integer
Attempt to join on values while ignoring leading zeroes.
Example: Values 001 and 1 are considered a match.
standardize - Decimal
Attempt to join on values and ignore leading/trailing zeroes.
Example: Values 001.00, 1, and 1.00 are considered a match.
standardize - String
Attempt to join on values while ignoring leading spaces, case, and any ASCII character other than A-Z, 0-9, +, or -.
Example: Values blue (with leading and/or trailing spaces), blue, and BLUE are considered a match.
- Click OK.
-
Run the job now or schedule the job to run later.
Note: If you discover a join and the total number of rows generated exceeds 2^32 (4294967296), some join metadata List Views may be unavailable.