LintData

A "linter" for pandas DataFrames to automate data quality audits.

Installation

You can install LintData via pip:

pip install lintdata

Via UV:

uv add lintdata

Or install from source:

git clone https://github.com/patelheet30/lintdata.git
cd lintdata
pip install -e .

Features

✅ 20+ Data Quality Checks - Missing values, duplicates, outliers, type consistency, and more
✅ Zero Configuration - Works out of the box with sensible defaults
✅ Highly Configurable - Customize thresholds and select specific checks
✅ Multiple Export Formats - Text, HTML, JSON, and CSV reports
✅ Custom Checks API - Extend with your own validation logic
✅ Pandas Native - Integrates seamlessly via .lint accessor

Quick Start

import pandas as pd
import lintdata

# Load your DataFrame
df = pd.read_csv("your_data.csv")

# Run quality checks
report = df.lint.report()
print(report)

Example Output:

--- LintData Quality Report ---
Shape: (1000, 8)

Running Checks:
Found 5 issue(s):
  1. [Missing Values] Column 'age': 45 missing values (4.5%)
  2. [Duplicates] Found 12 duplicate rows (1.2% of data)
  3. [Outliers] Column 'salary': 8 potential outliers detected (IQR method)
  4. [Mixed Types] Column 'phone' contains both numeric and string values
  5. [High Cardinality] Column 'user_id' has 987 unique values (98.7%)

--- End of Report ---

Available Checks

LintData includes 22+ built-in checks across multiple categories:

• Missing Data: Missing values, missing patterns
• Duplicates: Duplicate rows, duplicate columns
• Data Types: Mixed types, type consistency
• Statistical: Outliers, skewness, correlation warnings
• Categorical: Cardinality, rare categories, case consistency
• Numerical: Negative values, zero inflation
• Strings: Whitespace, special characters, length outliers
• Dates: Format consistency, future dates, date range anomalies
• Multi-table: Referential integrity (foreign key validation)

Export Formats

Save reports in multiple formats:

# HTML report with visualizations
df.lint.report(report_format='html', output='report.html')

# JSON for programmatic access
df.lint.report(report_format='json', output='report.json')

# CSV for spreadsheet analysis
df.lint.report(report_format='csv', output='issues.csv')

Custom Checks

Extend LintData with your own validation logic:

def check_email_format(df):
    """Validate email addresses."""
    warnings = []
    for col in df.select_dtypes(include='object').columns:
        if 'email' in col.lower():
            invalid = df[~df[col].str.contains('@', na=False)]
            if len(invalid) > 0:
                warnings.append(f"[Email] Column '{col}': {len(invalid)} invalid emails")
    return warnings

# Register and use
df.lint.register_check(check_email_format)
df.lint.report()

Documentation

Full documentation available at: LintData Documentation

Issues and Support

For general help or to report bugs, please open an issue on GitHub: LintData Issues.

If you have questions or need assistance, feel free to reach out via Discord: patelheet30