Attributes

Now, each entity is made up of attributes—these are the characteristics or properties of that entity. If “Customer” is the entity, its attributes might be Name, Email Address, Date of Birth, etc. Attributes help define what you know about an entity. In the same way that a customer has attributes that describe them, a product entity might have attributes like Price, Category, or Manufacturer. Attributes are essential because they give you the details that define your entities.

In Daana, attributes are a little more than what they usually are in an information model, the meta data you need to provide can be found when you "expand" the entity by pressing the "little right arrow" (>).

Column types in Daana

The column type in Daana is as such a little more than your typical data type also since it tells Daana how to treat the specific attribute when building the data pipelines later.

Start & End Timestamp

One might say, why not just have "Timestamp"? Well, in Data Warehousing it's often important to understand what a timestamp is signifying. If the concept in the attribute is transactional in nature, like for instance a credit card transaction, then you only need to use the "Start Timestamp" and can omit the "End Timestamp". But in cases where there is a clear start and end, for instance when describing the duration of a trip from A to B, or between when and when a customer had a status, the start & end timestamps becomes particularly powerful.

String Value

String values are typically exactly the same as corresponding concept in a database, i.e. if you can have characters in the data, like in a name, then it should be string.

Numerical Value

Numerical values are also similar to corresponding concept in a database, and should be only numerical values, no characters allowed.

Unit of Measure

Unit of Measure might feel like a special case, however this is also very powerful in a Data Warehouse setting. What is does that attributes numerical value actually represent? Dollars? Pieces? Hours? What? Knowing this is useful in the subsequent analytics you want to do on your transformed data.