📊 EDA Automation Tool (Streamlit)

An interactive Streamlit web application that automates Exploratory Data Analysis (EDA) for any uploaded CSV file.
It provides dataset diagnostics, missing value analysis, statistical summaries, and visual insights through univariate and bivariate analysis.

---

🚀 Features

✅ Dataset Overview

• Dataset shape (rows, columns)
• Data preview (top rows)
• Schema summary:
  • Column names
  • Data types
  • Missing value counts and percentages
• Smart warnings:
  • Columns with high missing values
  • High-cardinality categorical columns (ID-like features)

---

✅ Univariate Analysis

Numeric Columns

• Summary statistics: count, mean, median, std, skewness
• Histograms with KDE
• Boxplots for outlier visualization
• Skewness interpretation labels

Categorical Columns

• Unique count and mode
• Top-N category distributions
• Percent share + “Others” bucket
• High-cardinality warnings

---

✅ Correlation Analysis

• Pearson and Spearman correlation options
• Masked correlation heatmap (lower triangle)
• Top correlated feature pairs table (interactive Top-N)

---

✅ Bivariate Analysis

Numeric vs Numeric

• Scatter plot
• Correlation value

Categorical vs Numeric

• Grouped boxplot (numeric distribution across categories)
• Option to limit category count for readability

Categorical vs Categorical

• Crosstab table
• Stacked bar chart (count or percentage mode)

---

🧠 Tech Stack

• Python
• Streamlit
• Pandas / NumPy
• Matplotlib
• Seaborn

---

📂 Project Structure

eda-automation-tool/
│
├── main.py
├── requirements.txt
└── README.md

2️⃣ Install dependencies

pip install -r requirements.txt

3️⃣ Run the app

streamlit run main.py

---

📌 Use Cases

• Quick dataset understanding before ML modeling
• Missing value diagnostics & schema validation
• Feature relationship analysis
• Detecting outliers and class/category imbalance

👤 Author

Ishatva Singh Panwar