datakit-local 0.1.5

Modern web-based data analysis tool - process CSV/JSON/EXCEL/PARQUET files locally with SQL

DataKit

Modern web-based data analysis tool

Process CSV/JSON/XLSX/PARQUET files locally with complete privacy. No data ever leaves your machine.

Quick Start

# Install DataKit
pip install datakit-local

# Start DataKit (opens browser automatically)
datakit

# Or start server without opening browser
datakit serve --no-open

Features

• Complete Privacy: All data processing happens locally
• Large Files: Process CSV/JSON files up to 4-5GB
• Fast Analysis: DuckDB-powered SQL engine via WebAssembly
• Modern Interface: React-based web UI
• Visualizations: Built-in charts and data exploration
• Advanced Queries: Full SQL support with auto-completion

Installation

Requirements

• Python 3.8 or higher
• Modern web browser (Chrome, Firefox, Safari, Edge)

Install from PyPI

pip install datakit-local

Usage

Basic Commands

# Start DataKit (default behavior)
datakit

# Start server only
datakit serve

# Start and open browser explicitly  
datakit open

# Start on custom port
datakit serve --port 8080

# Start on custom host (network accessible)
datakit serve --host 0.0.0.0 --port 3000

# Start without opening browser
datakit serve --no-open

Information Commands

# Show version and features
datakit version

# Show system information
datakit info

# Check for updates
datakit update

Options

Option	Description	Default
-p, --port	Specify port number	Auto-detect (3000-3100)
-h, --host	Specify host address	127.0.0.1
--no-open	Don't open browser automatically	Opens browser
--reload	Enable auto-reload (development)	Disabled

🔧 Advanced Usage

Custom Configuration

from datakit import create_app, find_free_port
import uvicorn

# Create custom app
app = create_app()

# Find available port
port = find_free_port()

# Run with custom settings
uvicorn.run(app, host="0.0.0.0", port=port)

Programmatic Usage

import datakit

# Start server programmatically
datakit.run_server(host="localhost", port=3000)

Use Cases

Perfect for:

• Data Scientists: Analyze datasets without cloud dependencies
• Privacy-Conscious Users: Process sensitive data locally
• Enterprise Environments: No data leaves your network
• Large File Analysis: Handle multi-GB files efficiently
• SQL Analysis: Query your data with full SQL support

Security & Privacy

• Local Processing: All computation happens in your browser
• No Data Upload: Files never leave your machine
• No Internet Required: Works offline after installation
• Enterprise-Safe: Perfect for sensitive data analysis

Supported File Formats

• CSV: Comma-separated values with auto-detection
• JSON: Nested JSON files with flattening support
• Large Files: Optimized for files up to 4-5GB

Comparison with Other Tools

Feature	DataKit	Pandas	Excel	Cloud Tools
File Size Limit	Couple of GBs	Memory Limited	1M rows	Varies
Privacy	Complete	Complete	Complete	Limited
SQL Support	Full	Limited	None	Varies
Setup Time	1 command	Code required	Manual	Account setup
Browser Interface	✅	❌	❌	✅
Offline Use	✅	✅	✅	❌

Related Packages

• Node.js: npm install -g datakit-cli
• Docker: docker run -p 8080:80 datakit/app
• Homebrew: brew install datakit (coming soon)

Examples

Analyze Sales Data

# Start DataKit
datakit

# Upload your sales.csv file
# Write SQL queries like:
# SELECT product, SUM(revenue) FROM sales GROUP BY product
# Create visualizations with built-in charts

Process Large Datasets

# DataKit handles large files efficiently
datakit serve

# Load multi-GB files with streaming processing
# Query with pagination for smooth performance

License

AGPL-3.0-only License - see LICENSE file for details.

Support

• 📚 Documentation: https://docs.datakit.page
• 💬 Discussions: https://discord.gg/grKvFZHh
• 🌐 Website: https://datakit.page

Acknowledgments

Built with:

• FastAPI - Modern Python web framework
• Click - Command line interface
• DuckDB - High-performance analytical database
• React - User interface library

---

DataKit - Bringing powerful data analysis to your local environment with complete privacy and security.