When creating calculated columns in your Data Warehouse, you are asked to define a path describing how the table you are creating a column on is related to the table you are pulling information from. To successfully create a path, you need to know two things:
If you know this information, you can easily create a path following the instructions in this topic. You may want to ask a technical expert in your organization or contact the Professional Services team.
This concept is covered in the Understanding and evaluating table relationships article, but a quick summary never hurt anyone, right?
Tables can be related to one another in one of three ways:
Relationship Type |
Example |
---|---|
one-to-one |
The relationship between people and driver’s license numbers. A person can have only one driver’s license number, and a driver’s license number belongs to only one person. |
one-to-many |
The relationship between orders and items - an order can contain many items, but an item belongs to a single order. In this case, the orders table is the one side and the items table is the many side. |
many-to-many |
The relationship between products and categories: a product can belong to many categories, and a category can contain many products. |
When a relationship between two tables is understood, it can be used to determine what path should be created to bring information from one table to another. This next step requires knowing the primary and foreign keys that facilitate a table relationship.
A Primary Key
is an unchanging column or set of columns that produces unique values within a table. For example, when a customer makes an order on a website, a new row is added to the orders
table in your shopping cart, with a new order_id
. This order_id
allows both the customer and business to track the progress of that specific order. Because order id is unique, it is typically the Primary Key
of an orders
table.
A Foreign Key
is a column created inside a table that links to the Primary Key
column of another table. Foreign Keys create references between tables, allowing analysts to easily look up and link records together. Say you wanted to know which orders belonged to each of your customers. The customer id
column (Primary Key
of the customers
table) and the order_id
column (Foreign Key
in the customers
table, referencing the Primary Key
of the orders
table) allows us to link and analyze this information. When creating a path, you are asked to define both the Primary Key
and Foreign Key
.
When creating a column in your Data Warehouse, you must define the path that brings information from one table into another. Sometimes paths pre-populate because a path exists between tables, but if this does not happen, you must create one.
Use the relationship between customers and orders to show you how it is done. Broken down:
one-to-many
- a customer can have many orders, but an order can have only one customer. This tells us the direction of the relationship, or where the calculated column should be created. In this case, it means information from the orders
table can be brought into the customers
table.primary key
you want to use is customers.customerid
, or the customer ID
column in the customers
table.foreign key
you want to use is orders.customerid
, or the customer ID
column in the orders
table.Now, you can create the path.
Click Data > Data Warehouse.
In the table list, click the table in which you want to create the column. In this example, it is the customers
table.
The table schema displays. Click Create New Column.
Give your column a name, for example, Customer's orders
.
Select the definition for the column. Check out the Calculated Column Guide for a handy cheat sheet.
In the Select table and column dropdown, click the Create new path option.
Using the dropdowns, select the primary and foreign keys for each table.
On the Many
side, you select orders.customerid
- remember, customers can have many orders.
On the One
side, you select customers.customerid
- an order can only have one customer.
Click Save to save the path and finish creating the column.
Commerce Intelligence cannot guess primary/foreign key relationships. You do not want to introduce incorrect data into your account, so creating paths must be done manually.
Currently, paths can only be specified between two different tables. Does the logic that you are trying to recreate involve more than two tables? It then might make sense to (1) join the columns to an intermediary table first, then to the “final destination” table, or (2) consult with the Professional Services team to find the best approach to your goals.
A column can only be the foreign key reference for ONE path at one time. For example, if order_items.order_id
points to orders.id
, then order_items.order_id
cannot point to anything else.
Many-to-many
paths can technically be created, but often produce bad data because neither side is a true one-to-many
foreign key. The best way to approach these paths always depend on the specific desired analysis. Consult the RJ analyst team to uncovering the best solution.
If you are prevented from creating a calculated column due to one or more of the limitations above, contact support with a description of the column you are
Created an incorrect path in your Data Warehouse? Or maybe you are doing a little spring cleaning and want to tidy up? If you need to delete a path from your account, you can send a ticket over to Adobe support analysts. Be sure to include the name of the path!
Now that you are comfortable creating paths for calculated columns in your Data Warehouse. If you are still unsure about a particular path, remember that you can always click Support in your Commerce Intelligence account to get assistance.