Post

Cyber Shujaa Exploratory Data Analysis Assignment

Posted
By Grace Sampao
2 min read
Cyber Shujaa Exploratory Data Analysis Assignment

From Blueprints to Binaries: An Architectural Approach to Titanic EDA 🏛️ ➡️ 💻

In architecture, the integrity of a structure depends on the site investigation conducted before the first stone is laid. In Data Science, that investigation is Exploratory Data Analysis (EDA).

As an architect pivoting into AI, I have applied this structural mindset to my latest project: a deep dive into the Titanic dataset as part of the Cyber Shujaa Data and AI Specialist track.

The Project Foundation 📐

The primary goal of this assignment was to develop hands-on experience with the standard EDA workflow. Using a dataset of 891 passengers from the Pandas documentation, I explored 12 initial features to understand what factors influenced survival.

The Toolkit 🛠️

  • Language: Python.
  • Libraries: Pandas, Numpy, Matplotlib and the Math module.
  • Environment: Jupyter Notebook and VS Code.

Building the Model: Data Cleaning & Feature Engineering 🏗️

Just as we categorize building materials by their properties, I refined the raw data to make it “load-bearing” for analysis.

  • Handling Missing Values: Addressed gaps in the Cabin, Age and Embarked columns.
  • Data Typing: Converted Survived, Pclass, Sex and Embarked into categorical data types for better processing.
  • Efficient Replacement: Utilized np.where() to replace values, realizing that Numpy is essential for simplifying the EDA process.
  • Feature Engineering: Created three new features ⎯ Age Group, Accompanied and Cabin Assignment ⎯ bringing the total to 15 features.
  • Categorical Mapping: Transformed discrete age values into specific age group categories.

Visualizing the Structure 📊

One of my biggest “aha!” moments involved Matplotlib. I learned that it requires data in an array-like structure to plot effectively. This understanding allowed me to create a multivariate analysis of survivors and casualties based on whether they were travelling “accompanied”.

Accompanied vs Unaccompanied Survivors per Age Group

Structural Insights 💡

  • Class Disparity: Cabin allocation was tied directly to passenger class; First Class had the highest number of allocations, while Third Class had the lowest.


Cabin Allocation by Passenger Class



  • Port Statistics:

    • Port S recorded the highest fare revenue, while Port Q recorded the lowest.


Fare Revenue by Port


  • Port S recorded the lowest fares while Port C the highest.


Age vs Fare Paid per Port



  • Survival Rates: Unfortunately, the total number of casualties was higher than the number of survivors.


Survival Rates

The Roadmap Ahead 🚀

While I have successfully mastered changing column data types and basic plotting, my journey is just beginning:

  • Next Steps: I plan to conduct a more concise analysis of the Survived target variable.
  • Learning Goals: I am diving deeper into Numpy, exploring advanced Matplotlib customization and studying Data Structures as they apply to Pandas.

I am aiming to secure my first role in Data Science and AI by June this year, bringing a unique blend of analytical and design-oriented skills to the table.

Exploratory Data Analysis, Cyber Shujaa
cyber shujaa data science exploratory data analysis pandas matplotlib python
This post is licensed under CC BY 4.0 by the author.