CodeAlpha_DataAnalytics

Internship Projects

CodeAlpha — Data Analytics Internship Projects

Intern: Ədalət
Organization: CodeAlpha
Period: 2026
Tools: Python · pandas · matplotlib · seaborn · scikit-learn · BeautifulSoup · VADER · TextBlob

---

Project Overview

This repository contains all 4 Data Analytics tasks completed during the CodeAlpha internship program. Each task is self-contained with its own dataset, code, and visual outputs.

Task	Topic	Dataset	Key Output
Task 1	Web Scraping	books.toscrape.com	books_data.csv (1,000 books)
Task 2	Exploratory Data Analysis	books_data.csv	4 charts + statistical summary
Task 3	Data Visualization	books_data.csv	6 charts + dashboard
Task 4	Sentiment Analysis	Amazon Fine Food Reviews	7 charts · LR 83.8% accuracy

---

Repository Structure

CodeAlpha_DataAnalytics/
│
├── Task1_WebScraping/
│   ├── task1_web_scraping.ipynb
│   ├── books_data.csv
│   └── README.md
│
├── Task2_EDA/
│   ├── task2_eda.ipynb
│   ├── chart_rating_distribution.png
│   ├── chart_price_distribution.png
│   ├── chart_price_by_rating.png
│   ├── chart_availability.png
│   └── README.md
│
├── Task3_DataVisualization/
│   ├── task3_visualization.ipynb
│   ├── viz1_rating_bar.png
│   ├── viz2_price_kde.png
│   ├── viz3_avg_price_rating.png
│   ├── viz4_rating_by_price_bucket.png
│   ├── viz5_violin_price_rating.png
│   ├── viz6_dashboard.png
│   └── README.md
│
├── Task4_SentimentAnalysis/
│   ├── task4_sentiment_analysis.ipynb
│   ├── sentiment_results.csv
│   ├── sent1_vader_distribution.png
│   ├── sent2_vader_vs_textblob.png
│   ├── sent3_compound_distribution.png
│   ├── sent4_confusion_matrix.png
│   ├── sent5_polarity_subjectivity.png
│   ├── sent6_wordcloud.png
│   ├── sent7_model_comparison.png
│   └── README.md
│
└── README.md  ← (this file)

---

Key Results Summary

Task 1 — Web Scraping

• Scraped 1,000 books across 50 pages from books.toscrape.com
• Extracted: title, price (£), star rating, availability
• Output: clean CSV with 0 null values, 0 duplicates

Task 2 — EDA

• Price range: £10.00 – £59.99 | Mean: £35.07 | Median: £35.98
• Rating distribution: uniform across 1–5 stars (no platform bias)
• Price–Rating correlation: r = 0.028 (near-zero)
• All 1,000 books: In Stock

Task 3 — Data Visualization

• 6 professional charts: horizontal bar, KDE histogram, grouped bar, stacked bar, violin plot, dashboard
• Key insight: price does not predict rating across any price bucket

Task 4 — Sentiment Analysis

• Dataset: 10,000 Amazon Fine Food Reviews (balanced, 2,000 per star)
• VADER: 78.9% positive · 17.7% negative · 3.4% neutral
• VADER ↔ TextBlob agreement: 74.5%
• Logistic Regression accuracy: 83.8% (best model)
• Random Forest accuracy: 78.6%

---

How to Run

All notebooks are designed for Google Colab. No local setup required.

1. Open colab.research.google.com
2. Upload the .ipynb file for the task you want to run
3. Upload the required dataset (if any) to the Colab session
4. Run all cells sequentially (Runtime → Run all)

Dependencies

# Task 1–3
pip install requests beautifulsoup4 pandas matplotlib seaborn

# Task 4 (additional)
pip install vaderSentiment textblob wordcloud scikit-learn

---

Analytical Notes

• Correlation ≠ causation — all r values reported are observational only
• All ML models use stratified train/test split (80/20) with random_state=42 for reproducibility
• No data leakage — preprocessing fitted on training set only
• Balanced sampling used in Task 4 to prevent class bias

---

CodeAlpha Data Analytics Internship · 2026