DEV Community: Fariq

Power BI: Schemas and Data Modelling

Fariq — Sun, 01 Feb 2026 19:38:58 +0000

Introduction

In the generation of proper, efficient, and meaningful reports in Power BI, data modeling is essential. Within an educational context, e.g., the analysis of student performance, enrollment, or attendance data, a properly organized data model provides credible data, enhanced performance, and simplified report building. The important concepts of data modeling are schema, table relationships, fact tables, and dimension tables.

Star Schema

The simplest and most preferable data modelling tool used in Power BI is the star schema, as it is a simple and performance-friendly method. A fact table could hold quantifiable information in a model of education, like student exam scores, number of attendances, or course completion rates. This fact table is related to various dimension tables, including Student, Course, Teacher, School, and Date.
The student performance fact table, as an example, would have the following fields: studentID, courseID, dateID, and score, whereas the dimension tables will have the following: student name, grade level, course name, and term. Such organization enables teachers and administrators to easily examine performance based on subject, class, or year.

Snowflake Schema

A snowflake schema happens when the dimension tables are further subdivided into other related tables. An example is a school dimension that may be divided into county, sub-county, and school type separate tables. Although this decreases duplication of data, it adds additional relationships to the model.
Snowflake schemes may complicate and slow down education dashboards in Power BI, particularly in the process of developing DAX measures. Due to this reason, the star schema is commonly used to report enrollment trends or student performance unless the institution needs data normalization.

Relationships

Tables are related by means of relationships in a Power BI model. Relationships in an education dataset are normally one-to-many so that one student may be shown in a large number of exam records, or one course may even have a large number of students enrolled. The proper relationship direction means that one will have the filters, like choosing a given grade level or academic term, appropriately influence all the related visuals.
Improperly structured relationships may cause wrong averages, absence of student records, or wrong enrollment counts.

Fact and Dimension Tables

The quantitative education data in fact tables includes test scores, hours of attendance, or credits earned.
Dimension tables hold descriptive data such as student demographics, course names, school locations, and academic calendars.
The division of these tables enhances clarity and also enables the teachers, administrator, and analyst to investigate the data in various ways.

Relevance of Good Data Modeling in Education

Good data modeling in Power BI is necessary to:

Accurate student performance analysis
Efficient dashboard performance, even with large enrollment datasets
Simpler DAX calculations for KPIs such as pass rates and retention rates
Better decision-making by school administrators and education stakeholders

It can be structured in a way that allows the education institutions to monitor the trends, detect learning gaps, and facilitate data-driven enhancements in schools and programs when strong data modelling practices are implemented properly.

MS Excel for Data Analytics

Fariq — Sun, 25 Jan 2026 09:28:53 +0000

What is MS Excel?

Microsoft Excel is a spreadsheet application used to store, organize, and analyze data. It is widely used in schools, offices, and businesses because it is easy to learn and powerful enough for basic data analysis tasks.

Understanding the Excel Workspace

Excel data is stored in cells, which are arranged in rows (numbers) and columns (letters). A group of cells makes up a worksheet, and multiple worksheets form a workbook.

Entering and Organizing Data

You can type data such as names, numbers, or dates directly into cells. Excel allows you to format data using tables, bold text, colors, and borders to make it easy to read and analyze.

Sorting and Filtering Data

Excel helps you quickly organize data using Sort (arranging data in order, e.g., A–Z or smallest to largest) and Filter (showing only data that meets certain conditions).

Using Basic Formulas

Excel can perform calculations using formulas. Common beginner formulas include:

SUM – adds numbers
AVERAGE – finds the mean
COUNT – counts entries

Formulas always start with an equals sign (=).

Creating Charts and Graphs

Excel can turn data into visual charts like bar charts, line graphs, and pie charts. Charts help users easily understand trends and comparisons in data.

Why Excel is Useful for Data Analytics Beginners

Excel is ideal for beginners because it is:

Easy to use
Widely available
Good for basic data analysis
Visual and interactive

Conclusion

Microsoft Excel is a great starting tool for data analytics. By learning how to enter data, use simple formulas, sort and filter information, and create charts, beginners can easily analyze data and make informed decisions.

From "What Happened?" to "What Should We Do?" Understanding the Four Types of Data Analytics

Fariq — Tue, 20 Jan 2026 20:20:52 +0000

Every day, your applications generate mountains of data. User clicks, API calls, error logs, transaction records. It all piles up. But raw data sitting in a database is like having a library where all the books are written in a language you don't speak. The real value lies in asking the right questions and knowing which analytical approach will provide the answers you need.

Let me walk you through the four types of data analytics that transform noise into insight, each answering a progressively more sophisticated question about your data.

Descriptive analytics: the rearview mirror

The question: What happened?

Descriptive analytics is your starting point. It takes historical data and organizes it into a format that humans can understand. Dashboards, reports, charts. Think of it as looking in the rearview mirror while driving. You're not trying to predict where you're going. You're simply getting a clear picture of where you've been.

Real-life example:

Imagine you're running an e-commerce platform. Descriptive analytics tells you that last month you processed 15,000 orders, your average order value was $67, and mobile traffic increased by 23% compared to the previous quarter. Is your monitoring dashboard showing API response times over the past 24 hours? That's descriptive analytics. Your weekly active users chart? Descriptive analytics.

These insights don't explain why mobile traffic jumped or what caused that spike in API latency at 3 AM last Tuesday. They just show you the facts. But those facts are crucial. You can't diagnose problems or predict the future without first understanding your baseline reality.

In practice:

Most developers encounter descriptive analytics daily through tools like Grafana, Datadog, or Google Analytics. Is that graph showing your deployment frequency? Is your error rate trending over time? Your database query performance metrics? All descriptive. It's the foundation everything else is built on.

Diagnostic analytics: the detective work

The question: Why did it happen?

Once you know what happened, the natural next question is why. Diagnostic analytics digs into your data to find causes, correlations, and connections. It's detective work. Looking for clues that explain the patterns you observed.

Real-life example:

Let's say your descriptive analytics revealed that user signups dropped 40% last week. That's concerning, but knowing it happened isn't enough. Diagnostic analytics helps you investigate. Did the drop coincide with a recent deployment? Was there a change in your ad spending? Did a particular browser start having compatibility issues with your registration form?

Maybe you drill down and discover that 95% of the drop came from mobile users on iOS, and it started exactly when you deployed version 2.4.1. You check the logs and find a JavaScript error that only triggers on Safari. Mystery solved.

In practice:

This is where you use techniques like drill-downs, filters, and correlation analysis. You might join your application logs with deployment timestamps, compare conversion rates across different user segments, or analyze A/B test results to understand why variant B performed better. Tools like your ELK stack, Splunk, or even SQL queries with GROUP BY and WHERE clauses become your investigative toolkit.

Diagnostic analytics doesn't just identify problems. It also helps you understand successes so you can replicate them.

Predictive analytics: the crystal ball

The question: What will happen?

Now we're getting into the future. Predictive analytics uses your historical data, statistical models, and often machine learning to forecast what's likely to happen next. It's not magic. It's pattern recognition at scale.

Real-life example:

Your SaaS application has been collecting user behavior data for two years. Predictive analytics can help you identify which users are at high risk of churning based on patterns like decreased login frequency, reduced feature usage, or specific sequences of actions that historically preceded cancellations.

Or consider capacity planning. Based on historical growth trends, seasonal patterns, and current trajectory, predictive models can estimate that you'll need to scale your database cluster by 30% before the holiday shopping season. This gives you time to prepare instead of scrambling when traffic spikes.

In practice:

This is where machine learning enters the chat. You might build regression models to forecast server load, classification models to predict user churn, or time series models to anticipate seasonal traffic patterns. Libraries like scikit-learn, TensorFlow, or even simpler statistical approaches can power these predictions.

The key insight: predictive analytics doesn't tell you what will happen with certainty. It tells you what's likely to happen based on what you know. It's probabilistic, not deterministic. But that probability can be incredibly valuable for planning and preparation.

Prescriptive analytics: the strategic advisor

The question: What should we do about it?

This is the most sophisticated level. Prescriptive analytics takes everything you've learned (what happened, why it happened, what's likely to happen) and recommends specific actions to achieve your goals.

Real-life example:

Your predictive model says you're likely to see a 50% traffic increase next month. That's useful information, but what should you actually do about it? Prescriptive analytics might recommend:

Scaling your Kubernetes cluster to 15 pods (not 10, not 20, but exactly 15 based on cost performance optimization)
Increasing your database connection pool size by 40%
Shifting 30% more budget to your CDN
Implementing rate limiting for endpoints that become bottlenecks under load

Or imagine a recommendation engine. It doesn't just predict that a user might like sci-fi books. It prescribes showing them specific titles in a specific order at specific times to maximize engagement and conversion.

In practice:

Prescriptive analytics often combines optimization algorithms, simulation, and business rules. It might use techniques like linear programming to optimize resource allocation, reinforcement learning to recommend actions in dynamic environments, or decision trees to map out the best course of action given multiple constraints.

This is where data science truly becomes a decision-making tool. You're not just observing or predicting. You're actively shaping outcomes.

Bringing it all together

These four types of analytics aren't separate silos. They build on each other in a natural progression:

Describe what's happening in your systems and user base
Diagnose why those patterns exist
Predict what's likely to happen next
Prescribe what actions will move you toward your goals

As developers, we often start with descriptive analytics (monitoring, logging, metrics) because it's the most accessible. But the real power comes from climbing this ladder. Each level requires more sophisticated tooling and techniques, but each also delivers exponentially more value.

The next time you're looking at a dashboard or debugging an issue, ask yourself: Am I just describing what I see, or am I working toward prescribing what I should do? Understanding which question you're trying to answer will guide you to the right analytical approach and the right insights.

What level of analytics does your team currently use? The data is already there, waiting to tell you not just what happened, but what you should do next.

Git Made Simple

Fariq — Fri, 16 Jan 2026 10:40:43 +0000

It is normal for developers to save different versions of their work, track changes, make changes, and collaborate; this is the work of Git.

What is Git

Git is a version control system that lets one:

Track changes
Go to earlier code versions
Collaborate
Save code online on GitHub

Steps

Step 1: Installing Git Bash

Go to Git Bash

Select your Operating System, e.g., Windows, macOS, or Linux

Install Bash

Open Git Bash

You will see a terminal that looks like this:

username@computer MINGW64 ~

Step 2: Create (or Log Into) a GitHub Account

Go to Github
Create an account or log in if you already have one

Step 3: Connect Git Bash to GitHub

Set up your name

git config --global user.name "Your Name"

Set up your email

git config --global user.email "youremail@example.com"

Note: Make sure the email matches your GitHub email.

Check Your Configuration:

git config --list

Step 4: Create a New Project or Repository(repo)

Create a Folder

mkdir my-first-git-project

Navigate to the now-created folder

cd my-first-git-project

Initialize Git

git init

Now Git is watching this created project

Step 5: Track Changes with Git

Create a File

touch index.html

Open the Folder Directly from Git Bash

Input this in Git Bash and press enter:

code .

This opens VS Code: The dot (.) means “this current folder."

Step 6: Create the First code in VS Code

Open index.html and add this code:

<!DOCTYPE html>
<html>
<head>
  <title>My First Git Project</title>
</head>
<body>
  <h1>Hello World!</h1>
</body>
</html>

Step 6: Check Git Status in Git Bash

Return to Git Bash and run:

git status

Git will show index.html as untracked.

This means Git sees the file, but it has not started tracking it yet.

Step 7: Add the File to Git Tracking

Track the index.html by running:

git add index.html

Or add everything using:

git add .

Step 8: Commit Your Changes

git commit -m "Initial HTML file commit"

The m means a message to be added to the commit

Step 9: Create a Repository on GitHub (Online)

Now that your project exists locally on your computer, the next step is to create an online repository on GitHub where your code will be stored and shared.

Make sure you are logged in to Github

On the right top corner of the github website, click the + icon.

Select New repository

Enter a Repository name; use the same name as your folder i.e:

my-first-git-project

Choose Public

Scroll to the bottom and click Create repository

Github will now show some commands

Step 10: Connect Your Local Project to GitHub

Return to Git Bash and run:

git remote add origin https://github.com/YOUR-USERNAME/my-first-git-project.git

Replace YOUR-USERNAME with your actual GitHub username

Step 11: Rename the Default Branch**

Run:

git branch -M main

Step 12: Push Your Code to GitHub

Run:

git push -u origin main

Note:

git push - send your code online
origin - the GitHub repository
main - the branch name
-u - remembers this destination for future pushes

Step 13: Go back to your GitHub repository page and Refresh the page

You will now see the index.html file and the commit message

Step 14: Pull Code from GitHub

This will let the latest code be pulled or downloaded back to your local machine, where you can again edit the code. Run:

git pull origin main

DEV Community: Fariq

Power BI: Schemas and Data Modelling

Introduction

Star Schema

Snowflake Schema

Relationships

Fact and Dimension Tables

Relevance of Good Data Modeling in Education

MS Excel for Data Analytics

What is MS Excel?

Understanding the Excel Workspace

Entering and Organizing Data

Sorting and Filtering Data

Using Basic Formulas

Creating Charts and Graphs

Why Excel is Useful for Data Analytics Beginners

Conclusion

From "What Happened?" to "What Should We Do?" Understanding the Four Types of Data Analytics

Descriptive analytics: the rearview mirror

Diagnostic analytics: the detective work

Predictive analytics: the crystal ball

Prescriptive analytics: the strategic advisor

Bringing it all together

Git Made Simple

What is Git

Steps

Step 1: Installing Git Bash

Step 2: Create (or Log Into) a GitHub Account

Step 3: Connect Git Bash to GitHub

Step 4: Create a New Project or Repository(repo)

Step 5: Track Changes with Git

Step 6: Create the First code in VS Code

Step 6: Check Git Status in Git Bash

Step 7: Add the File to Git Tracking

Step 8: Commit Your Changes

Step 9: Create a Repository on GitHub (Online)

Step 10: Connect Your Local Project to GitHub

Step 11: Rename the Default Branch**

Step 12: Push Your Code to GitHub

Note:

Step 13: Go back to your GitHub repository page and Refresh the page

Step 14: Pull Code from GitHub

Summarized Git Workflow