Roll-Up or Drill-Down, which has the most Detail in Data Warehouse?

Question

Context

Studying on data Warehouse(DW) I came across the content regarding the types of OLAP servers and the types of operations that can be done, within this context I was confused regarding the concept of detailing about two operations, they are called Roll-Up and Drill-Down

In the book the author states that the operation of Drill-Down has greater detailing but the example given did not make me understand well.

Exemplification of the Book.

Chapter 29 - Overview of Data Warehousing and OLAP. Below is how the operations are visually exemplified by the author of the book.

Roll-Up Example

---

Drill-Down Example

What the Teradatapoint website Claims

As it was not very clear the explanation, I went to Google and found this teradatapoint article that made me understand better.

About the Roll-Up Operation

My Understanding

The content illustrated on the site teradatapoint, made me better understand the subject and if I could summarize or exemplify in a few words the operations, it would be more or less like this:

Roll-Up Operation Search reduce the 3D cube for a more streamlined analysis, for example: Perform a bimetral analysis of a given product in a given region.

---

Drill-Down operation looks for augment the 3D cube, for an analysis with a longer time interval, for example perform analyses at annual intervals of a given product in a given region.

I understand by detailing the data, an increasingly smaller delimitation of the scope involved, which indicates a decrease in the size of the cube, but the book states the opposite.

Doubt

I didn’t understand in which of the two operations there is a larger data detailing, the book claims it’s the Drill-Down and the therapist’s website says it’s the Roll-Up.

Source: Database Systems 6th edition, authors: Elmasri, Ramez Navathe, Shamkant B. Year: 2011 Cap 29 pg 724 e 725

Answer 1

Well, I’m new around here and this is my first answer. As my knowledge of the subject is also not the most profound, I will try to be as brief as possible.

OLAP

When the database is built on OLAP, it means that it is no longer relational, meaning there is no increase or decrease in the amount of data, they are already in a format whose recovery is usually faster than in a relational database. However, there are versions of OLAP, ranging from ROLAP (with a relational engine), MOLAP (with a multidimensional engine) and an intermediate HOLAP (Hybrid). Unfortunately I’ve never worked with a hybrid to give any further explanation.

As for Drill-Down and Roll-Up

When accessing information from a OLAP (i.e., Multidimensional) database, it is possible that it has two or more dimensions. Some of these dimensions may have hierarchies. The highest hierarchies are the aggregated information of the lower hierarchies, for example: One side of the dimension can be constructed with information from the location of customers. In this case, Brazil would be the highest hierarchy. When applying a Drill-down, this hierarchy can be divided in several ways, such as regions (North, South, Southeast etc.) or states (SP, RJ etc.). The lower you go in the detail of the information, the more likely you are to find cells without measurements in the same way the information becomes detailed.

As for Roll-Up, you can think about the product dimension because it is easier to understand. You may be organizing the database of a retail store with various products. So each product would be the lowest cell in the hierarchy of this dimension. Thus, when applying the aggregation of sales of a given product group (Liquificador -> Kitchen Products -> Household Appliances, for example), the complication existing at this point is the fact that the information is not standardized among the products, for example, which will require some form of data transformation.

I hope I answered. Hugs!