dataguard-cli 0.1.3

A fast, modular CLI tool for data quality analysis, profiling, and cleaning using Polars and Rich.

DataGuard CLI

DataGuard CLI is a fast, modular data quality tool that helps developers validate, analyze, and clean datasets before using them in analytics or machine learning pipelines.

Built with Polars, Rich, and Typer, it provides a powerful command-line interface for profiling datasets and detecting issues such as missing values, duplicates, and inconsistent formats.

---

🚀 Features

• ⚡ Fast data profiling using Polars
• 🧠 Rule-based issue detection engine
• 📊 Dataset quality scoring system
• 🔍 Detailed column-level analysis
• 🛠️ Auto-fix functionality for cleaning data
• 📁 Batch processing for multiple datasets
• 🔒 Strict mode for pipeline enforcement
• 🎨 Rich CLI interface with tables and panels
• ⚙️ Configurable validation rules (YAML-based)
• ✅ Fully testable and modular architecture

---

📦 Installation

From PyPI

pip install dataguard-cli

From Source

git clone https://github.com/your-username/dataguard-cli.git
cd dataguard-cli
pip install -e .

---

🧑‍💻 Usage

Scan a dataset

dataguard scan data.csv

Show only important issues

dataguard scan data.csv --summary

Export report

dataguard scan data.csv --export report.json

Strict mode (CI/CD pipelines)

dataguard scan data.csv --strict

---

Fix dataset issues

dataguard fix data.csv
dataguard fix data.csv --apply-changes

---

Generate structured report

dataguard report data.csv
dataguard report data.csv --export report.json

---

Batch processing

dataguard batch ./datasets/

---

📊 Example Output

📊 Dataset Quality Score: 78/100 🟠

Column   Type     Null %   Severity   Issues
--------------------------------------------------------
age      Int64    45%      3          High missing values
email    String   5%       2          Invalid format
salary   Float64  0%       0          Clean

---

⚙️ Configuration

Edit rules in:

dataguard/config/rules.yaml

Example:

null_thresholds:
  critical: 30
  warning: 0

duplicate_threshold: 0
constant_column: true
email_validation: true

---

🧪 Testing

pytest

---

🏗️ Architecture

CLI → Commands → Core Engine → Rules → UI

• CLI: Typer-based command interface
• Core: Rule engine and processing logic
• UI: Rich-based rendering
• Config: YAML-driven rules

---

💡 Why DataGuard?

Data quality issues are a major cause of:

• incorrect analytics
• unreliable ML models
• pipeline failures

DataGuard provides a lightweight and extensible way to enforce data validation before downstream usage.

---

🔮 Future Improvements

• Schema validation support
• ML-based anomaly detection
• Integration with data pipelines (Airflow, etc.)
• Plugin-based rule system

---

🛠️ Tech Stack

• Python
• Polars
• Rich
• Typer
• Pytest

---

👨‍💻 Author

Sarthak Dongare (Frosty-8)