Understanding Dimensional Modeling

×
Master the Art of Data Science: A Step-by-Step Guide to Becoming an Industry-Ready Data Scientist
Download Roadmap

Home

Tanmay Maheshwari — Published On February 28, 2023

Advanced Data Engineering Data Warehouse Database Technique

Scan Objects in Full Color in 5 Minutes or

Less with True Portability & Unrivaled LEARN MORE
Accuracy

Introduction
One of the most important assets of any organization is the data it produces on a daily basis. This data is used by an
organization to find valuable insights which help in improving an organization’s growth and strategies and give them an
upper hand over its competitors. This article explains to you the idea of Dimensional Modeling as part of Data
Warehousing with different steps involved.

Learning Objective

Understand Data Warehousing

Understand the difference between Normalised and Denormalised form
Learn Dimensional Modeling and its implementation using real-world application
But this data, both current and historical, can only be utilized to its full potential when it is easily accessible and
available. So now the question arises of how we can store this data in such a way that it fulfills the above requirements.
Here comes the concept of Data Warehousing.

Table of Contents
1. Data Warehousing
2. Dimensional Modeling
3. Four Steps of Designing a Dimensional Model

Data Warehousing
Data warehousing is a technique in which information is stored in a central repository that can be used by business
analysts, data engineers, data scientists, and decision-makers through business intelligence (BI) tools to make decisive
decisions.
It is used to handle large amounts of data to understand relationships and trends across the data.
Advantages of Data Warehousing –

1. It makes information easily accessible.

2. It presents information consistently.
3. It adapts to changes in the data.
4. It presents information in a timely way.
5. It is secure, so it protects the information assets.
We use cookies on Analytics Vidhya websites to deliver our services, analyze web traffic, and improve your experience on the site. By using Analytics Vidhya, you

Data in Data Warehouse is stored in aagree

tabular format
to our Privacywhich canTerms
Policy and be normalized or denormalized.
of Use. Accept
 Difference Between
Understanding Normalized
Dimensional and Denormalized Form
Modeling
In normalized form, data is stored in multiple tables, reducing data redundancy and inconsistency, thus achieving data
integrity. In the denormalized form, data is stored in a limited number of tables (maybe a single table) to reduce
querying time.
Both of them contain joined tables, but the key difference between them is the degree of normalization. As the degree
of normalization increases, the complexity of the model increases, and as the complexity of the model increases, the
time to retrieve data also increases.

Dimensional Modeling
The data model used to store data in the denormalized form is called Dimensional Modeling. It is the technique of
storing data in a Data Warehouse in such a way that enables fast query performance and easy access to its business
users. It involves creating a set of dimensional tables that are designed to support business intelligence and reporting
needs.
The goal of dimensional modeling is to provide a simple and intuitive way to access and analyze data, making it easy for
business users to understand and use it. It aims at making simple data models. When the data models are as simple as
possible, they can be understood easily, allowing the software to navigate and deliver results quickly and efficiently.
The core concept of dimensional modeling is the creation of a star schema. It is called so as the tables are arranged in
the form of a star. Dimensional modeling includes facts and dimensions. Let’s have a basic idea of what Facts and
Dimensions are.
Fact tables contain measures or numerical data associated with a business process, like the number of products sold. In
contrast, dimensional tables store the description or textual information related to the business process, like who
bought the products. We will discuss facts and dimensions in detail later in this article.
A dimensional model represents the different business processes of an organization. A fact table with its dimension
table is a single business process.
Each dimensional model consists of many fact tables, with each fact table joined with corresponding dimension tables. A
fact table is connected to another fact table via a common dimensional table between them; this common dimensional
table is called a bridge table. We can even connect a fact table with a fact table directly, but it is not a wise option as it
makes the model complex and difficult to understand.

Source: Javatpoint

A dimensional table is connected to the fact table using the foreign key in the fact table. The dimensional table is the
parent table, and the fact table is the child table.
A dimensional model contains the same information as a normalized model. Still, the data in a dimensional model is
We use cookies on Analytics Vidhya websites to deliver our services, analyze web traffic, and improve your experience on the site. By using Analytics Vidhya, you
packed in such a way that delivers user understandability, query performance, and resilience to change.
agree to our Privacy Policy and Terms of Use. Accept
 Implementation
Understandingof Dimensional
Dimensional Modeling in Real World
Modeling
Dimensional Modeling can be best understood by implementing it in real-world cases. We will pick up the E-Commerce
industry like Myntra, Flipkart, Amazon, etc., as it is familiar to everyone. We will be creating a dimensional model
depicting different business processes that take place in an E-Commerce industry.

Four Steps of Designing a Dimensional Model

Step 1: Select the Business Process
The first step involves selecting the business process, and it should be an action resulting in output.
Business Process #1:The e-Commere industry is widely known for selling and buying goods over the internet, so our
first business process will be the products bought by the customers.
Business Process #2: Delivery status is also one of the most important business processes in this industry. It tells us
where the product is currently from. It’s dispatched from the warehouse to the customer’s given address.
Business Process #3: Maintaining the inventory in order to ensure that items don’t run out of stock, how sales are going
on etc.

Step 2: Decide the Grain of each Business Process

A grain is a business process at a specified level. It tells us what exactly a row, in fact, a table, represents. All the rows in a
fact table should result from the same grain. Each fact table is the result of a different grain selected in a business
process. The grain should be as granular (at the lowest level) as possible.
Grains for the above business processes are
Grain 1: We can have the grain as the products purchased by the customer, i.e., each row of the fact table will represent
all the products checked out by the customer from the cart but suppose a customer ordered 100 products, so this will
be represented as a single row. Imagine how complex it will become to query such data, so we must choose a grain as
granular as possible. Therefore, our grain will be an individual product ordered by a customer, i.e., one product per row.
This will make the data simple and easy to query.
Similarly, we will select the most granular grains for the remaining processes.
Grain 2: Here also, the grain will be the status of an individual product shipped from the warehouse to the delivery
location.
Grain 3: Here, each row will represent the daily inventory for each product in each store., it will tell the stock of that
product left in the inventory and how many products have already been sold.

Step 3: Identify the Dimensions for the Dimensional Table

Before identifying the dimensions we will understand what a dimensional table is.

Dimensional Tables
These are the tables that are joined to fact tables. It describes the “who, what, where, when, how, and why” associated
with the business event. It contains the descriptive attributes used for grouping and filtering the facts.
Some important points regarding Dimension Tables:

1. It stores textual information related to a business process.

2. It answers the ‘who, what, where, when, why, and how’ questions related to a particular business process.
3. Dimension tables have more columns and less number of rows.
4. Each dimension table has a primary key that is joined to its given fact table.
5. Dimension attributes are the primary source of query constraints, grouping, and filtering.

We use cookies on Analytics Vidhya websites to deliver our services, analyze web traffic, and improve your experience on the site. By using Analytics Vidhya, you

agree to our Privacy Policy and Terms of Use. Accept

 Dimensions describe
Understanding the measurements
Dimensional Modeling of the fact table. For example, customer id is a measurement, but we can
describe its attributes further, more as what is the name of the customer, the address of the customer, gender, etc.
Our dimensional model will have the following dimensions:

Date Dimension: This dimension table is used in almost every dimensional model as it helps monitor the business’s
performance with time.

Product Dimension: This table will contain information regarding the product ordered.

Order Dimension: This detail will contain information regarding the order.

Customer Dimension: This dimension table will contain the customer’s information

Promotion Dimension: This table covers the promotion condition under which the product was sold. The promotion
conditions include temporary sales, reduction in price, discounts, etc.

Warehouse Dimension: This table will contain information about the different warehouses located across the country.
We use cookies on Analytics Vidhya websites to deliver our services, analyze web traffic, and improve your experience on the site. By using Analytics Vidhya, you

agree to our Privacy Policy and Terms of Use. Accept

 Understanding Dimensional Modeling

Step 4: Identify the Facts for the Dimensional Table

This is the final step in which we have to decide which facts (measurements) must be included in the fact table, but
before that, let’s discuss what a fact table is.

Fact Table

The term fact represents a business measure; therefore, a fact table in dimensional modeling stores the performance
measurements resulting from a business process. These performance measurements measure the business, i.e., these
are the metrics through which we can infer whether our business is in profit or loss. Different business measurements
can be unit price, number of goods sold, etc. Each row in a fact table is a business event that results in measurements,
and each fact table represents a business process in the organization. Now the event depends upon the grain we select.
The selection of grain plays a vital role in the success of our dimensional model as it helps in selecting the measurements
in the fact table to which further dimension tables are joined.
Since we have chosen three business processes, we will have three fact tables, but sometimes we get confused about
whether an attribute should be added to the fact table or dimension table. To avoid that confusion, we will be using the
following points to identify whether an attribute is a fact or dimension:

1. Textual data is generally stored in dimension tables where, whereas numeric data is generally stored in the fact
table.
2. Continuous valued numeric values are stored in the fact tables, whereas discrete numeric values are stored in the
dimension table.
3. The values that constantly change are kept in the fact table, whereas values that remain static or change very less
with time are kept in the dimensional table.

Fact Table 1:

Grain: Individual product of the order per row.

So we will select the measurements corresponding to this grain. When we check out, the measurements that come are
unit price, quantity, ordered, discount, etc., so we have added these measurements.
In this way, we add our measurements to a fact table. Weight also could have been added as a measurement, but its
value remains constant therefore, we will keep it in the dimension table.
Fact Table 2:
We use cookies on Analytics Vidhya websites to deliver our services, analyze web traffic, and improve your experience on the site. By using Analytics Vidhya, you

agree to our Privacy Policy and Terms of Use. Accept

 Understanding Dimensional Modeling

Grain: Delivery Status of individual products in the order

This fact table will tell us the delivery status of the product, i.e., the location of the product it got delivered, etc.
Fact Table 3:

Grain: Daily inventory for each product in each store.

So through this fact table, we will track the stock of the different products. This is how we create our dimensional model
by following the above steps.
Our final model will look like this:

There are three types of fact tables:

1. Transaction Fact Tables: It records a row in the table whenever there is a transaction. Here the transaction is the
grain itself. The first fact table in our model is a transaction fact table in which different transactions between
customers and companies are depicted.
2. Periodic Fact Tables: These record a row in the table for a definite period of time. That period of time can be daily,
monthly, etc. Here the grain is the period of time we select. Example – The inventory fact table that we have taken is
an example of a periodic fact table in which we will be calculating the stock of the products on a daily basis.
We use 3. Accumulating
cookies Fact websites
on Analytics Vidhya Tables: toItdeliver
stores ourpredictable steps
services, analyze between
web traffic, the process’s
and improve beginning
your experience andByend.
on the site. usingWhenever a you
Analytics Vidhya,

agree to
predictable step is recorded, the table is our Privacy and
revisited Policyupdated.
and TermsUpdation
of Use. Accept
of a row is unique to this type of fact table
 only as compared
Understanding to the other
Dimensional types of fact tables. Example: The order’s delivery status is an example of the
Modeling
accumulating fact table in which we have to update the product’s location whenever it reaches the desired location
until it is delivered.
4. Factless Fact Tables: It records a row in the table with no numeric measurement, but that grain is important to
store. This table is also used to record data that didn’t happen. Example: A table showing the products on sale which
were not sold.

Some important points regarding Fact Tables:

1. A fact table generally contains numeric and additive facts. E.g., the number of products sold, the cost of each
product, etc
2. Facts can also be semi-additive or non-additive.
3. Generally, textual data is not stored in a fact table; if required, it can be stored.
4. Continuous numeric values are stored in the fact table. E.g., The cost of products can take value in a broad range.
5. We should not store redundant textual information in fact tables unless the text is unique for every row in the fact
table.
6. We should avoid using 0 for no activity in the fact table.
7. Fact tables have more rows and less number of columns.
8. Fact tables have two or more foreign keys that connect to the dimension table’s primary keys.
9. It answers the question, “What is the business process measuring.”

Conclusion
I hope through this article; you got a basic understanding of Dimensional modeling. This article shows how fact and
dimension tables are created and the steps we follow to implement a dimensional model successfully.

Key Takeaways

Data Warehouse is a central repository used to store data that data analysts and data engineers can use for analysis
purposes.
Dimensional modeling is a technique used to store normalized forms of data.
Dimensional modeling consists of creating Fact and Dimension tables.

data warehouse Database normalization Dimensional modeling fact table

About the Author

Tanmay Maheshwari

Our Top Authors

We use cookies on Analytics Vidhya websites to deliver our services, analyze web traffic, and improve your experience on the site. By using Analytics Vidhya, you

agree to our Privacy Policy and Terms of Use. Accept

 Understanding Dimensional Modeling
view more

Download
Analytics Vidhya App for the Latest blog/Article

Next Post
Step-by-Step Roadmap to Learn SQL in 2023

Leave a Reply
Your email address will not be published. Required fields are marked *

Comment

Name*

Email*

Website

Notify me of follow-up comments by email. Notify me of new posts by email.

Submit

We use cookies on Analytics Vidhya websites to deliver our services, analyze web traffic, and improve your experience on the site. By using Analytics Vidhya, you

agree to our Privacy Policy and Terms of Use. Accept

 Understanding Dimensional Modeling

Top Resources

10 Best AI Image Generator Tools to Use in 2023

Nitika Sharma - AUG 17, 2023

© Copyright 2013-2023 Analytics Vidhya. Privacy Policy Terms of Use Refund Policy

We use cookies on Analytics Vidhya websites to deliver our services, analyze web traffic, and improve your experience on the site. By using Analytics Vidhya, you

agree to our Privacy Policy and Terms of Use. Accept