DEV Community: Max Njiru

WHAT IS DATA MODELLING IN POWER BI ?

Max Njiru — Mon, 14 Sep 2026 01:10:48 +0000

Data Modelling

Data modelling in Power BI is the process of organizing data into tables and defining how those tables are connected so that Power BI can efficiently analyze and report on the data.
In a real world business environment, data is rarely stored in one perfect table. Instead, information is often distributed across several tables. For example, a retail company may have separate tables for:

Customers
Products
Sales Transactions
Dates
Locations
Employees

Data modelling determines how these tables should be organized and how they should communicate with each other.
For example, a business may have a FactSales table containing sales transactions and separate tables such as DimCustomer, DimProduct,_ DimDate_ and DimLocation.
Dim simply means Dimension
In Power BI, a dimension table is a table that gives descriptive information about the things you are analyzing

The purpose of the model is not simply to connect tables. The goal is to create a structure that allows Power BI to correctly answer business questions such as:

How much did the company sell?
Which products generated the highest revenue?
Which customers purchased the most?
Which month had the highest sales?
Which location generated the highest profit?
How has sales performance changed over time? Therefore, data modelling is the foundation that allows Power BI to turn separate pieces of raw data into one meaningful analytical system.

Why is a Well-Designed Data Model Important?

A well designed data model is extremely important because it directly affects the accuracy, performance, usability, scalability, and maintainability of a Power BI report.
A beautiful dashboard cannot compensate for a poorly designed data model. If the underlying tables and relationships are incorrect, visualizations may display incorrect numbers even if the charts look professional.

Importance for Reporting

The purpose of Power BI is to create report that helps organizations understand their businesses. A well designed data make it possible to combine information from different tables without manually putting everything in one table
suppose a company wants to create a report showing:
Total Sales by Product Category and Month
The sales amount maybe stored in:
factsales
while the product category is stored in:
Dimproduct
and the month is stored in:
Dimdate
The relationship between these tables allows Power BI to bring the information together.
Without a proper model, the analyst may have to manually combine data before every analysis.
Therefore:
A good data model makes reporting easier because users can select fields from different tables and analyze them together.

Importance for Analytics

Data modelling is also important for deeper business analysis.
Businesses often need to analyze relationships between different areas of their operations.

Importance for DAX Calculations

DAX is the formula language used in Power BI to create measures, calculated columns, and other calculations.
For example:
Total sales =Sum(FactSales{Sales Amount})
This simple measure calculates the total sales stored in the fact table.
However, the real power comes when the measure is used with dimension tables.
A good data model makes DAX calculations more reliable, reusable, and easier to understand.
A poorly designed model can result in:

Incorrect totals
Unexpected filter behavior
Complicated DAX formulas
Duplicate calculations
Ambiguous relationships

Importance for Performance

Data modelling also has a major effect on Power BI performance.
Performance refers to how quickly Power BI can:

Load data
Refresh data
Calculate measures
Apply filters
Generate visualizations
Respond to user interactions

Importance of scalability

Scalability means the ability of a data model to grow without requiring a complete redesign.
A business may start with:

Sales
Customers
Products Later, it may want to add: Dates
Locations
Employees
Suppliers
Promotions
Sales Channels

Importance of Maintainability

Maintainability refers to how easy it is to understand, update, troubleshoot, and modify a data model over time.
A Power BI report is rarely created once and never changed. Businesses continuously introduce:

New products
New customers
New locations
New transactions
New business rules
New reporting requirements A clear model makes these changes easier to manage. For example: DimCustomer- DimProduct- DimDate- DimLocation- FactSales A developer can quickly understand what each table represents. If a problem occurs with customer information, the developer knows where to look: DimCustomer If there is an issue with sales transactions: FactSales If there is a problem with dates: DimDate This is much easier than trying to find an error inside one enormous table containing hundreds of columns. Therefore: A well organized model reduces maintenance effort and makes it easier for different developers or analysts to work on the same Power BI solution.

Importance for Data Accuracy

Data modelling is also important for ensuring that reports produce correct results.
Relationships determine how filters move between tables.

Importance for Reducing Data Redundancy

Data redundancy means storing the same information repeatedly when it does not need to be repeated.
For example, imagine 10,000 sales records where the product category Electronics is repeated in every row.
Instead of storing the same descriptive information repeatedly, a dimensional model can store it once in DimProduct

Importance of Business Decision Making

Ultimately, the purpose of data modelling is to support better business decisions.
A well designed model allows managers and analysts to move from raw data to useful information.
E.g;
Raw Data-Data Modelling-Relationships-DAX calculations-Power BI Reports-Business Insights-Better Decisions
A company can use these insights to determine:

Which products should be promoted?
Which locations are underperforming?
Which customers generate the most revenue?
Which months have the highest demand?
Where should inventory be increased?
Which products are producing low profits? Therefore, data modelling is a critical foundation for turning data into actionable business intelligence.

Summary

Data modelling in Power BI is the process of organizing tables, defining relationships, and creating a logical structure that allows data to be analyzed effectively.The key principle is that a Power BI report is only as strong as the data model underneath it. A well designed model creates a reliable foundation for relationships, DAX calculations, visualizations, analytics, and future expansion.

GETTING STARTED WITH EXCEL

Max Njiru — Sun, 30 Aug 2026 05:54:33 +0000

Introduction

Microsoft Excel is a spreadsheet tool that organizes data into rows and columns. Excel is useful in data cleaning, filtering and analysis.
In this article, we will start from basic excel capabilities and gradually build up to advanced data cleaning.

Core Concepts of Excel

Ribbon in Excel

Ribbon is the menu bar you see at the top of Excel sheet
It contains different tools and commands that help you work with your work

1.Home — basic formatting, alignment, number formats, and editing tools.
2.Insert — used to insert tables, PivotTables, charts, and other visual elements.
3.Page Layout **— controls how worksheets are arranged and displayed when printed.
4.Formulas **— provides access to Excel functions and formula-related tools.

Data **— used for data analytics. It contains tools for sorting, filtering, removing duplicates, data validation, and other data management activities. 6.Review — includes tools for reviewing, commenting, and protecting worksheets. 7.View **— controls how the workbook and worksheets are displayed.

Rows,Columns,Cells

Rows- runs horizontally across the worksheet and are usually represented by numbers.
Columns - runs vertically across the worksheet and are represented by letters
Cell - Formed when a row and a column meet e.g D7

Data Types

A data type is basically the different information contained in a cell. It can take different forms e.g Text, Numbers, Currency, Percentage

Data Cleaning

I learnt some basic data cleaning techniques such as;

Removing duplicates
Handling missing values
Removing extra spaces with_ TRIM() 4. Standardizing text with UPPER(), _LOWER(),PROPER()
Fixing inconsistent dates

Raw Uncleaned data

Cleaned data

Data Validation

The restrictions on the type of data user can input into a cell. (Dropdown lists, Numbers , ranges, date limits, text length)

Data Cleaning And Formatting

Removing duplicates: Prevents inflated results in analysis.
Mombasa
Mombasa
Standardizing text this ensures consistency
NAIROBI
NairoBi
checking dates: corrects inconsistent format and invalid entries, fixing words format:
use PROPER() to standardize capitalization in names.

Conclusion

In conclusion data cleaning in excel is more the a technical step, it is the foundation of reliable analysis. This is a effective data cleaning turns Excel from a simple spreadsheet tool into a powerful engine for analytics.

GIT WORKFLOW

Max Njiru — Mon, 24 Aug 2026 18:04:18 +0000

During my practical sessions at Luxdev HQ, I started working with Git and GitHub as part of my development workflow. At first, some of the commands seemed to do almost the same thing, especially git add, git commit, and git push. With practice, I started understanding that each command represents a different stage of the workflow
In this article, I will explain each stage using a simple Python project and some of the commands I have been practicing.

1. Starting With a Simple Project

For this example, imagine that I am working on a small Python calculator project.

My project folder might look like this:
mycalculator/

calculator.py
README.md

Inside calculator.py, I could start with something very simple

print("Welcome to my calculator")
At this point, I am just working normally on my computer. I can open the file in VS Code, change the code, save it, and continue working.

This is where the working directory comes in

2. Understanding the Working Directory

`The working directory is basically the project folder where I am doing my actual work.

If I create a new file, edit an existing file, or delete a file, those changes happen in my working directory.

print("Welcome to my Python calculator")
I have changed the file, but I have not yet told Git that I want to record this change.

This is where I found git status very useful.
I started thinking of git status as my what is going on? command.

Whenever I am not sure what Git knows about my project, I can run git status and see which files have been changed, which files are untracked, and which changes are staged.

3. Understanding Untracked and Modified Files

One thing a beginner can easily find confusing is the difference between an untracked file and a modified file.

Suppose I create a new file called:
hello.py
Git may show it as an untracked file
On the other hand, if Git already knows about calculator.py and I change something inside it, Git can tell me that the file has been modified.
So I can think about it like this:
Untracked
→ Git has not started tracking this file.

Modified
→ Git already knows the file, but I have changed it.

4. The Staging Area

This was one of the parts I had to think about the most when learning Git.

At first, I wondered:

If I have already changed my file, why do I need to add it?

I eventually understood that Git gives me a staging area where I can select the changes that I want to include in my next commit.

For example:
git add calculator.py
This tells Git that I want the changes in calculator.py to be included in my next commit.

The important thing I learned is that git add does not upload the file to GitHub.

11. The Difference Between Commit and Push

It simply moves the change from my working directory into the staging area.

5. Creating a Commit

Once I am satisfied with the changes I have staged, I can create a commit.
git commit -m "Update calculator message"