DEV Community: Stephen Omengo

Understanding Data Distributions and Their Impact on Data Science

Stephen Omengo — Tue, 23 Jun 2026 08:35:07 +0000

INTRODUCTION

Data is the foundation of every data science project, but raw data alone is of little value unless we understand how it behaves. The distribution of data is one of the most fundamental concepts in statistics and data science. Data distributions describe how values are spread across a dataset, providing important information about trends, patterns, and anomalies.

What is a data distribution?
This is a representation of how frequently different values occur within a dataset. It provides insight into the shape, center, spread, and variability of the data.

By analyzing distributions, data scientists can answer questions such as:

-Are the data points clustered around a central value?
-Is the data skewed toward higher or lower values?
-Are there unusual observations or outliers?
-Which statistical methods are appropriate for the data?

Visual tools such as histograms, density plots, and box plots are commonly used to examine distributions.

Common Types of Data Distributions

(a). Normal Distribution

Often called the Bell Curve, it is one of the most important distributions in statistics.

Examples:

-Human heights
-IQ scores
-Measurement errors

Many machine learning algorithms and statistical techniques assume that data follows a normal distribution

(b). Uniform Distribution

In Uniform Distribution, all outcomes have an equal probability of occurring.

Examples:

-Rolling a fair die
-Random number generation

They are frequently used in simulations and probability modeling.

(c). Binomial Distribution
This models the number of successes in a fixed number of independent trials.

Examples:

-Coin tosses
-Email campaign conversions
-Product quality inspections

(d). Poisson Distribution
The Poisson Distribution measures how often events occur within a fixed interval.

Examples:

-Website visits per minute
-Customer arrivals at a store
-Network failures

(e). Exponential Distribution
The Exponential Distribution measures the time between events.

Examples:

-Time between customer purchases
-Time between system failures
-Waiting time for service requests

This distribution is commonly used in reliability engineering and operational analytics.

Why Data Distributions Matter in Data Science

Data distributions help data scientists understand data patterns, detect outliers, and select appropriate analytical methods. They influence data preprocessing, model performance, and statistical testing. By understanding how data is distributed, organizations can make more accurate predictions, identify risks, and support better decision-making.

Real-World Example: E-Commerce Analytics

Consider an online retailer analyzing customer spending patterns.

If spending follows a normal distribution:

-Average spending accurately represents most customers.

If spending is heavily right-skewed:

-A small group of customers contributes a large portion of revenue.
-The median becomes a more reliable measure than the mean.

This insight influences marketing strategies, customer segmentation, and revenue forecasting.

Conclusion

Data distributions are a cornerstone of data science. They provide valuable insights into how data behaves and influence nearly every stage of the analytical process, from data exploration and feature engineering to model development and business decision-making.

Understanding Subqueries and CTEs in SQL: A Complete Guide

Stephen Omengo — Tue, 23 Jun 2026 07:48:08 +0000

Working with relational databases often requires breaking down complex problems into manageable parts. Two powerful tools that help achieve this in SQL are subqueries and Common Table Expressions (CTEs). While they may seem similar at first, they serve different purposes and are best used in different scenarios.

This article explores what subqueries and CTEs are, their types, use cases, and how they compare in terms of performance and readability.

What is a Subquery?

A subquery is a query nested inside another SQL query. It is used to perform operations that depend on the result of another query.

Basic Example

SELECT name
FROM employees
WHERE salary > (
    SELECT AVG(salary)
    FROM employees
);

In this example:

The inner query calculates the average salary.
The outer query retrieves employees earning above that average.

👉 In simple terms, a subquery provides intermediate results to the main query.

Types of Subqueries

Subqueries can be categorized based on how they are used and how they interact with the outer query.

Single-row Subquery

Returns only one row.

SELECT name
FROM employees
WHERE department_id = (
    SELECT id FROM departments WHERE name = 'Sales'
);

Multi-row Subquery

Returns multiple rows and is used with operators like IN, ANY, or ALL.

SELECT name
FROM employees
WHERE department_id IN (
    SELECT id FROM departments WHERE location = 'Nairobi'
);

Correlated Subquery

Depends on the outer query and is executed once for each row.

SELECT name
FROM employees e
WHERE salary > (
    SELECT AVG(salary)
    FROM employees
    WHERE department_id = e.department_id
);

👉 This is more dynamic but can be slower due to repeated execution.

Nested Subquery

A subquery inside another subquery.

SELECT name
FROM employees
WHERE department_id = (
    SELECT id
    FROM departments
    WHERE location = (
        SELECT location
        FROM offices
        WHERE city = 'Nairobi'
    )
);

When Should Subqueries Be Used?

Subqueries are ideal when:

You need a value derived from another query
The logic is simple and contained
You want to filter results dynamically
You’re working with aggregates (AVG, MAX, MIN, etc.)

However, they can become inefficient or hard to read when deeply nested or correlated.

What are CTEs (Common Table Expressions)?

A Common Table Expression (CTE) is a temporary result set defined at the beginning of a query using the WITH keyword. It improves readability and organization, especially in complex queries.

Basic Example
WITH avg_salary AS (
    SELECT AVG(salary) AS avg_sal
    FROM employees
)
SELECT name
FROM employees, avg_salary
WHERE salary > avg_sal;

👉 Think of a CTE as a temporary named query you can reference within your main query.

Types and Use Cases of CTEs

Non-Recursive CTE

The most common type, used for simplifying complex queries.

WITH department_totals AS (
    SELECT department_id, SUM(salary) AS total_salary
    FROM employees
    GROUP BY department_id
)
SELECT *
FROM department_totals
WHERE total_salary > 50000;

Use case:

Breaking down large queries into readable parts

Recursive CTE

Used to handle hierarchical or tree-structured data.

WITH RECURSIVE employee_hierarchy AS (
    SELECT id, name, manager_id
    FROM employees
    WHERE manager_id IS NULL

    UNION ALL

    SELECT e.id, e.name, e.manager_id
    FROM employees e
    INNER JOIN employee_hierarchy eh
    ON e.manager_id = eh.id
)
SELECT * FROM employee_hierarchy;

Use case:

Organizational charts
Category hierarchies
Graph traversal
When Should CTEs Be Used?

CTEs are best when:

Queries are complex and need structure
You want to reuse a result multiple times
You need recursive logic
You want to improve readability and maintainability
Subqueries vs CTEs: A Clear Comparison

Readability Subqueries: Can become difficult to read when nested CTEs: Much cleaner and easier to understand

👉 Winner: CTEs

Performance Subqueries: Correlated subqueries can be slow Often re-executed multiple times CTEs: Sometimes optimized better by the database But in some systems, they may not be cached and can behave like inline views

👉 Winner: Depends on the database engine

For repeated logic → CTEs often better
For simple tasks → subqueries are fine

Reusability Subqueries: Cannot be reused easily CTEs: Can be referenced multiple times in the same query

👉 Winner: CTEs

Complexity Handling Subqueries: Good for simple conditions CTEs: Ideal for complex, multi-step logic

👉 Winner: CTEs

Recursion Subqueries: Cannot handle recursion CTEs: Support recursive queries

👉 Winner: CTEs

When to Use Each
Use Subqueries when:
The query is simple and short
You only need the result once
You’re filtering using aggregates
Use CTEs when:
The query is complex or layered
You need better readability
You want to reuse logic
You’re working with hierarchical data
Conclusion

Both subqueries and CTEs are essential tools in SQL, and understanding when to use each can significantly improve your queries.

Subqueries are concise and useful for straightforward operations
CTEs provide structure, clarity, and power for more advanced scenarios

In practice, experienced developers often prefer CTEs for maintainability, especially in large projects—but subqueries still have their place for quick, simple tasks.

Introduction to Python for Data Analytics

Stephen Omengo — Fri, 08 May 2026 18:57:15 +0000

Python is one of the most popular programming languages used in data analytics. It is easy, flexible, and has a rich ecosystem of libraries, making it perfect for processing, analyzing, and visualizing data. Organizations in finance, healthcare, marketing, technology, and more use Python to turn raw data into actionable insights that enable better decision-making.

What is Data Analytics? - This involves collecting, cleaning, organizing, and interpreting data to identify patterns and trends. Python simplifies these processes through powerful tools and frameworks designed specifically for data manipulation and analysis.

Python is highly recommended for data analytics. Here are some of the examples :

1.Big Ecosystem of Libraries

Python provides specialized libraries that simplify data analysis tasks.

They include:

Library	Objective
`NumPy`	Numerical computations
`Pandas`	Data manipulation and analysis
`Matplotlib`	Data visualization
`Seaborn`	Statistical visualization
`SciPy`	Scientific computing
`Scikit-learn`	Machine learning

These libraries help reduce development time by using pre-built functions instead of writing everything from scratch.

2.Easy to read and understand

Python has a clean syntax that is readable and resembles natural language. It is more user-friendly than many other programming languages for beginners.

Example :

sales = [1200, 1500, 1800, 2000]
average = sum(sales) / len(sales)

print("Average Sales:", average)

This simplicity allows analysts to spend more time working on problems and less time on complex programming structures.

Core Principles of Python in Data Analytics

Variables and Data Types Variables store data values in memory.

Example:

name = "Stephen"
age = 21
salary = 250000.00

Common Data Types :

Integers (int)
Floats (float)
Lists (list)
Dictionaries (dict)
Strings (str)

Lists

Lists are used to store multiple values.

Example:

fruits = ["Apple", "Banana", "Orange", "Mango"]

print(fruits)

Lists are useful when handling datasets containing many records.

Statements

Conditional statements allow decision-making within programs.

Example:

score = 85

if score >= 80:
    print("Excellent")
else:
    print("Needs Improvement")

This helps analysts classify and filter data.

Loops

Loops automate repetitive tasks.

Example:

sales = [100, 200, 300]

for item in sales:
    print(item)

Loops are commonly used to process rows of data efficiently.

Pandas

Pandas is one of the most important Python libraries for data analytics. It introduces two powerful data structures:

Series - one-dimensional labeled array that can store:

Numbers
Text
Boolean values
Dates

DataFrames - two-dimensional table made of rows and columns.
It is similar to:

Excel spreadsheets
SQL tables
CSV files

A DataFrame resembles a spreadsheet or SQL table. Example :

import pandas as pd

data = {
    "Name": ["Stephen", "Brian", "Charles"],
    "Age": [21, 25, 22]
}

df = pd.DataFrame(data)

print(df)

Output

      Name  Age
0    Stephen   21
1    Brian   25
2  Charles   22

Pandas allows analysts to:

Load datasets
Clean missing values
Filter records
Group data
Perform calculations

Analyzing Data Files

Python can read data from multiple sources such as CSV, Excel, JSON, and databases.

Example of reading a CSV file:

import pandas as pd

df = pd.read_csv("sales.csv")

print(df.head())

The head() function displays the first five rows of the dataset.

Data Cleaning

Real-world datasets often contain errors, duplicates, or missing values. Data cleaning improves data quality before analysis.

Example:

df.dropna(inplace=True)

This removes rows containing missing values.
Data cleaning is a critical step because inaccurate data leads to unreliable analysis results.

Data Visualization

Visualization helps analysts understand patterns and communicate findings effectively.

Example:

import matplotlib.pyplot as plt

months = ["Jan", "Feb", "Mar"]
sales = [1200, 1500, 1700]

plt.plot(months, sales)
plt.title("Monthly Sales")
plt.xlabel("Months")
plt.ylabel("Sales")
plt.show()

Common chart types include:

Line charts
Bar charts
Pie charts
Histograms
Scatter plots

Visualizations make complex data easier to interpret.

Basic Statistical Analysis

Python can perform statistical operations quickly.

Example:

import statistics

numbers = [10, 20, 30, 40, 50]

print("Mean:", statistics.mean(numbers))
print("Median:", statistics.median(numbers))

These statistical calculations help summarize datasets.

Python in Real-World Data Analytics

Business Analyst

Companies analyze sales performance, customer behavior, and market trends.

Financial Analytics

Banks and financial institutions use Python for fraud detection and investment analysis.

Social Media Analytics

Organizations analyze user engagement and sentiment from online platforms.

Advantages of Python in Data Analytics	Disadvantages of Python in Data Analytics
Easy to learn and read due to simple syntax	Slower execution speed compared to compiled languages like C++
Large collection of libraries such as Pandas, NumPy, and Matplotlib	Higher memory consumption
Strong community support and extensive documentation	Mobile development support is limited
Open-source and free to use	Multi-threading limitations because of the Global Interpreter Lock (GIL)
Excellent for data visualization and reporting	Can become slower when processing extremely large datasets
Supports automation of repetitive tasks	Dynamic typing may lead to runtime errors
Integrates easily with databases and cloud platforms	Dependency management can sometimes be challenging
Suitable for machine learning and artificial intelligence	Some libraries require additional optimization for high performance
Cross-platform compatibility (Windows, Linux, macOS)	Not ideal for highly performance-critical applications
Rapid development and faster prototyping	Complex environments may require many external packages

Conclusion

Python has revolutionized the world of data analytics by offering simple yet powerful tools to work with data. It is easy to read, has a lot of libraries, and is supported by a strong community. It is suitable for beginners and professionals alike. insights.

Python for data analytics learning provides opportunities across a wide range of industries and serves as a foundation for more advanced fields like machine learning and artificial intelligence.

Introduction to SQL: DDL, DML, and Data Querying

Stephen Omengo — Tue, 14 Apr 2026 08:23:46 +0000

Structured Query Language (SQL) is used to manage and manipulate databases. In this assignment, I learned how to build and interact with a database from scratch.

DDL vs DML

DDL (Data Definition Language) is used to define the structure of a database. Commands like CREATE, ALTER, and DROP are used to create tables and modify them.

DML (Data Manipulation Language) is used to work with the data inside the database. Commands such as INSERT, UPDATE, and DELETE help in adding, modifying, and removing data.

Practical Usage

I used CREATE to build tables such as students, subjects, and exam_results. I then used INSERT to populate them with data. UPDATE helped correct mistakes, and DELETE removed unwanted records.

Filtering Data

The WHERE clause is used to filter records. I used operators such as:

= (equals)
> (greater than)
BETWEEN (range filtering)
IN (multiple values)
LIKE (pattern matching)

These made it easy to retrieve specific data from the database.

CASE WHEN

CASE WHEN allows us to transform data into meaningful categories. For example, I used it to classify student performance into Distinction, Merit, Pass, and Fail.

Reflection

This assignment was both challenging and interesting. Understanding relationships between tables and writing correct queries required attention to detail. However, it improved my confidence in working with databases and SQL queries.

IOS 26.4 Beta : WHAT USERS NEED TO KNOW

Stephen Omengo — Sun, 12 Apr 2026 18:51:57 +0000

Key Features

New emojis
New Accessibility options
CarPlay AI Integration
Performance and stability improvements

New Emojis

. The new iOS 26.4 beta expands the user's library by adding new characters for messaging and social media, such as the Distorted face, a fan favourite.

New Accessibility options

. To users who have a strong visual effect, there is a new feature for them called "Reduce Bright Effects." This reduces intense flashing when interacting with the keyboard.

Carplay AI Integration

. With Apple CarPlay, users can now use third-party apps such as ChatGPT and Gemini through voice command.

Performance and Stability Improvements

. Like all the previous updates, this is a general improvement that fixes bugs, provides smoother performance, "Not laggy anymore."

Compatibility

. IOS 26.4 is available on iPhone 11 series and later, plus iPhone SE (2nd generation).

Summary

. Although iOS 26.4 may appear to be a minor update, it delivers meaningful improvements in performance, accessibility, and AI integration. For both everyday users and developers, it represents another step toward a more intelligent and optimized Apple ecosystem.

Understanding Data Modeling in Power BI: Joins, Relationships, and Schemas Explained.

Stephen Omengo — Tue, 07 Apr 2026 07:14:51 +0000

Data modeling is a critical step in building effective Power BI reports. It shows how data is structured, connected, and used to generate understanding. A well-designed model improves performance and ensures accuracy, while a poor one can lead to slow dashboards and incorrect results.
This article explains data modeling in Power BI, covering SQL joins, relationships, schemas, and how to implement each step in practice.

Data Modeling

Data modeling is the process of organizing data into tables and showing how those tables relate to one another. This happens after data is imported and transformed.
Purpose of data modelling:

Efficient calculations
Structure of data
improve efficiency and clarity

SQL joins

What are joins? They are used to combine rows from two or more tables based on a related column.

1.Inner join

Returns only matching rows from both tables.

2.LEFT JOIN (LEFT OUTER)

Returns all rows from the left table and matching rows from the right.

3.RIGHT JOIN (RIGHT OUTER)

Returns all rows from the right table and matching rows from the left.

4.FULL OUTER JOIN

Returns all rows from both tables.

6. LEFT ANTI JOIN

Returns rows from the left table with no match in the right.

7.RIGHT ANTI JOIN

Returns rows from the right table with no match in the left.

How to Create Joins in Power BI
.Open Power BI Desktop
.Click Transform Data
.Select a table
.Click Merge Queries
.Select the second table
.Choose a matching column
Select join type:
.Inner
.Left Outer
.Right Outer
.Full Outer
.Left Anti
.Right Anti
Click OK
.Expand merged columns

Relationships in Power BI

Type of relationships:
1.One-to-Many (1) -Most common type.
2.Many-to-Many (M) -Both tables contain duplicate values.
3.One-to-One (1:1) -Each record matches exactly one record in another table.

Creating relationships in Power BI

.Go to Home → Manage Relationships
.Click New
.Select tables and columns
.Set cardinality and filter direction
.Click OK

Conclusion

Data modeling in Power BI is the foundation of effective reporting. Understanding joins, relationships, and schemas allows you to build models that are both efficient and reliable.

Guide to Publishing and Embedding Power BI Reports on the Web with IFrames

Stephen Omengo — Tue, 07 Apr 2026 07:13:36 +0000

Introduction

In today’s data-driven world, organizations rely heavily on tools that transform raw data into valuable insights. One of the most powerful solutions for this purpose is Microsoft Power BI. It enables analysts and businesses to create interactive dashboards and reports that support informed decision-making. One of the most powerful tools for this purpose is Microsoft Power BI. It allows analysts and businesses to create interactive dashboards and reports to support decision-making.

This article provides a step-by-step guide on how to :

Create a Power BI workspace
Upload and publish a report
Generate an Embed code
Embed the report into a website using an iframe

What is Power BI? This is a business analytics tool developed by Microsoft that enables users to transform, clean data, build interactive visualizations, and connect to multiple data sources*Excel*.

What is Power BI Publishing? Publishing a Power BI report involves moving your report from Power BI Desktop to the Power BI Service cloud platform.

Prerequisites

Key things to consider before you start:

Power BI Desktop installed
A Power BI account
A complete report

Step 1: Create a Workspace in Power BI Service

A workspace is a shared environment where you store datasets, reports, and dashboards.

Tutorial

Go to Power BI Service: https://app.powerbi.com
Log in using your account
On the left sidebar, click Workspaces
Click + New workspace
Enter: .Workspace name .Description
Click Save

Step 2: Prepare Your Report

You should make sure that your report is: Clean and well-structured, has meaningful visuals, uses proper relationships, and DAX measures.

Step 3: Publish the Report to Power BI Service

Tutorial

Open your report in Power BI

Click File → Publish → Publish to Power BI
Sign in if prompted
Select your workspace created earlier
Click Select

Once uploaded, you’ll see a confirmation message.

Step 4: Verify the Report in Power BI Service

Head back to Power BI Service
Open your workspace
Locate your report under Reports
Click to open and confirm everything is working

Step 5: Generate the Embed Code

Power BI allows you to embed reports using Publish to Web or secure embedding.
Note: Publishing to the web makes it easily accessible, and anyone with the link can view it.

Tutorial

Open your report in Power BI Service
Click File → Embed report → Publish to web

Click Create embed code

Confirm the warning
Copy the generated iframe code

Step 6: Understanding the IFrame Code

The embed code will look like this:
<iframe width="800" height="600" src="https://app.powerbi.com/view?r=..." frameborder="0" allowFullScreen="true"></iframe>

Key Attributes:
.Width & height: Control size
.Src: Contains report URL
.Frameborder: Removes border
.AllowFullScreen: Enables full-screen mode

Step 7: Embed the Report into a Website

.You can embed this iframe into:

HTML websites
Blogs (Dev.to, Medium)
Web apps

Step 8: Test Your Embedded Report

After embedding:
.Open your website in a browser
. Check: Report loads correctly, Filters and visuals work, and Responsiveness.

conclusion

Publishing and embedding Power BI reports is an essential skill for any data analyst. It allows you to transform your work into a shareable and interactive experience accessible from anywhere.