Python Mini Project – Exploratory Data Analysis (EDA) &
Visualization
DataSet :
[Link]
H9yXq1HYBVaxoR?usp=sharing
Note: Choose Dataset Wisely
Same Dataset for Excel, PowerBI, SQL and Python Project
Submission Link: [Link]
Project Objective
To explore, understand, and visually analyse a real-world dataset using Python by applying
exploratory data analysis techniques and appropriate data visualisation methods.
Project Instructions
1. You will be provided with a dataset in CSV or Excel format.
2. Perform the analysis using Python (Jupyter Notebook preferred).
3. Use only EDA and visualisation techniques.
Do not build machine learning models.
4. The analysis must be aligned with the actual structure of the dataset.
5. If a specific column type is not available, adapt the analysis and clearly mention it.
6. Save the notebook as:
YourName_Python_EDA_Project.ipynb
Task 1: Import Libraries and Load Data
● Import required libraries:
○ pandas
○ numpy
○ matplotlib
○ seaborn
● Load the dataset into a pandas DataFrame.
● Display the first few rows.
What to show:
● Code cell loading the data
● Output showing sample rows
Task 2: Dataset Overview
● Display:
○ Shape of the dataset
○ Column names
○ Data types
● Use .info() and .describe() to summarise the data.
What to show:
● Output of dataset structure and statistical summary
● Short markdown note explaining the dataset size and types of variables
Task 3: Data Cleaning (EDA Level)
Perform basic cleaning required for analysis:
● Check for missing values and document them
● Handle duplicates if present
● Correct incorrect data types
● Standardise column names if necessary
What to show:
● Code and output for missing value and duplicate checks
● Brief explanation of cleaning decisions
Task 4: Univariate Analysis
Analyse individual variables based on their data type:
● Numerical variables
○ Histograms
○ Boxplots
● Categorical variables
○ Value counts
○ Bar plots
What to show:
● At least two plots
● Short interpretation for each plot
Task 5: Bivariate Analysis
Analyse relationships between variables:
● Numerical vs Numerical
○ Scatter plot
○ Correlation heatmap
● Categorical vs Numerical
○ Boxplot or bar chart
Select variable pairs based on availability in the dataset.
What to show:
● At least two bivariate plots
● Explanation of observed patterns or relationships
Task 6: Multivariate Analysis
● Use one of the following (as applicable):
○ Pairplot
○ Facet-based plots
○ Grouped bar charts
● Analyse interaction between more than two variables.
What to show:
● One multivariate visual
● Short explanation of insights
Task 7: Data Distribution & Outliers
● Identify skewness and spread of numerical variables
● Detect outliers using boxplots or IQR method
What to show:
● Visual evidence of distribution or outliers
● Interpretation of findings
Task 8: Visualization Summary
Create a final visual summary:
● 3–4 key plots that best describe the dataset
● Ensure plots have:
○ Titles
○ Axis labels
○ Proper scaling
Task 9: Key Insights
Write 3–5 concise insights derived from the EDA and visualisations.
Examples:
● Patterns in distributions
● Strong or weak relationships between variables
● Presence of outliers or anomalies
● Differences across categories
Submission Requirements
● Jupyter Notebook (.ipynb)
● Well-commented code
● Clear markdown explanations
● Visualisations embedded in the notebook
Evaluation Criteria
● Correct and logical EDA flow
● Appropriate visual selection
● Alignment with dataset structure
● Quality of interpretation
● Code readability and organisation