speedyeda 0.2.0

🚀 One-line data exploration for developers & data scientists

🚀 SpeedyEDA - Production-Ready Data Exploration

Professional-grade exploratory data analysis in one command.

Stop writing boilerplate! SpeedyEDA gives you complete exploratory data analysis in seconds — now with advanced statistical methods, data quality alerts, and interactive HTML reports that rival industry-standard tools.

---

💡 Love this project? ⭐ Star us on GitHub!

Your star helps others discover SpeedyEDA and motivates us to keep improving! 🙏

---

🆕 What's New in v0.2.0

✨ Advanced Statistical Analysis

• Multiple correlation methods (Pearson, Spearman, Kendall)
• Enhanced outlier detection (IQR + Z-score)
• Detailed quantile statistics (5th, 95th percentiles)
• Normality tests (Shapiro-Wilk)

🚨 Automated Data Quality Alerts

• Multicollinearity detection
• High cardinality warnings
• Duplicate row detection
• Class imbalance analysis
• Excessive missing value alerts
• Mixed data type detection

📊 Interactive HTML Reports

• Beautiful Plotly visualizations
• Click-to-zoom charts
• Standalone HTML files
• Professional styling

📊 SpeedyEDA vs The Competition

Feature	SpeedyEDA	ydata-profiling	Sweetviz	D-Tale
Basic Statistics	✅	✅	✅	✅
Multiple Correlations	✅ Pearson/Spearman/Kendall	✅	✅	✅
Outlier Detection	✅ IQR + Z-score	✅	✅	⚠️
Data Quality Alerts	✅ 8 types	✅	⚠️ Limited	❌
Interactive HTML	✅ Plotly	✅	✅	✅ Flask
Dataset Comparison	🔜 v0.3.0	❌	✅	❌
Target Analysis	🔜 v0.3.0	✅	✅	⚠️
Speed (10K rows)	⚡ <1s	~10s	~5s	~3s
Installation Size	📦 ~100MB	~500MB	~200MB	~300MB
One-Line CLI	✅	❌	❌	❌
Fun Mode	✅ 🎉	❌	❌	❌

Bottom Line: SpeedyEDA combines the speed of simple tools with the features of professional ones, plus a delightful UX.

---

✨ Core Features

• 📊 Automatic Statistics - Mean, median, mode, std, skewness, kurtosis, detailed quantiles
• 🔍 Advanced Missing Value Analysis - Patterns, correlations, recommendations
• 📈 Auto Visualizations - Histograms, boxplots, correlation heatmaps (static + interactive)
• 🔗 Multiple Correlation Methods - Pearson (linear), Spearman (monotonic), Kendall (ordinal)
• 🎯 Smart Outlier Detection - IQR method + Z-score with configurable thresholds
• 🚨 Data Quality Alerts - Multicollinearity, high cardinality, duplicates, class imbalance, and more
• 🎨 Beautiful Terminal Output - Colorful, emoji-rich displays using rich
• 📄 Interactive HTML Reports - Professional Plotly-based reports with click-to-zoom
• 🔧 Smart Presets - Pre-configured for ecommerce, surveys, finance
• 🔌 Plugin System - Extend with custom visualizations and metrics
• 🤝 Interactive Mode - Guided column and plot selection
• 📦 Batch Processing - Analyze multiple datasets at once

🚀 Quick Start

Installation

pip install speedyeda

Basic Usage

# Full analysis with data quality alerts
fasteda sales.csv --fun

# Generate interactive HTML report
fasteda data.csv --html report.html

# Use preset with plots and HTML
fasteda products.csv --preset ecommerce --plots --html ecommerce_report.html

# Interactive mode
fasteda survey.xlsx --interactive

# Batch processing with HTML reports
fasteda file1.csv file2.csv file3.csv --batch --html

# Disable advanced features for speed
fasteda huge_dataset.csv --no-advanced

Python API

import pandas as pd
from fasteda import analyze, save_report

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

# Full analysis with advanced features
results = analyze(df, fun=True, advanced=True)

# Check data quality alerts
if results['quality_alerts']:
    for alert in results['quality_alerts']:
        print(alert.message)

# Multiple correlation methods
correlations = results['advanced_correlations']
print(correlations['spearman'])  # Spearman correlation matrix

# Outlier detection
outliers = results['outliers']
for col, info in outliers.items():
    print(f"{col}: {info['count']} outliers ({info['percentage']:.1f}%)")

# Save detailed report
save_report(results, "sales_report.json")

📋 CLI Options

Flag	Description
--fun	🎉 Emojis and colorful output (highly recommended!)
--html <file>	📊 NEW! Generate interactive HTML report
--no-advanced	⚡ Disable advanced features for faster processing
--summary	📝 Plain text summary with insights
--plots	📊 Generate and save static visualizations
--save <file>	💾 Export report (JSON/TXT)
--interactive	🤝 Interactive column/plot selection
--preset <name>	🎯 Use preset (ecommerce, survey, finance)
--columns <cols>	🎯 Analyze specific columns only
--batch	📦 Process multiple files
--quiet	🤫 Suppress terminal output

🎯 Smart Presets

SpeedyEDA includes built-in presets tailored for common scenarios:

• 📦 ecommerce - Product analysis, sales trends, customer behavior
• 📋 survey - Response distributions, sentiment analysis, demographics
• 💰 finance - Time series, correlations, risk metrics
• 🔧 general - Comprehensive all-purpose exploration

fasteda sales.csv --preset ecommerce --plots --fun

🔌 Extend with Plugins

Build custom analysis functions:

from fasteda.plugins import register_plugin

@register_plugin("outlier_detection")
def detect_outliers(df, threshold=1.5):
    # Your custom analysis
    return results

📦 Supported Formats

• 📄 CSV (.csv)
• 📊 Excel (.xlsx, .xls)
• 🗂️ JSON (.json)
• ⚡ Parquet (.parquet)

🌟 Why SpeedyEDA?

Before SpeedyEDA:

import pandas as pd
import matplotlib.pyplot as plt
import seaborn as sns

df = pd.read_csv("data.csv")
print(df.describe())
print(df.info())
print(df.isnull().sum())
plt.figure(figsize=(10,6))
# ... 20+ more lines of boilerplate ...

With SpeedyEDA:

fasteda data.csv --fun

✨ One command. Complete analysis. Beautiful output.

🤝 Contributing

We'd love your help making SpeedyEDA even better!

• 🐛 Found a bug? Open an issue
• 💡 Have an idea? Start a discussion
• 🎨 Want to contribute? Submit a PR
• ⭐ Love SpeedyEDA? Star the repo!

📄 License

MIT License - see LICENSE file for details.

---

Made with ❤️ by Dawaman

If SpeedyEDA saves you time, ⭐ star the repo to show your support!

🐛 Report Bug · 💡 Request Feature · 📖 Documentation