PyEgen

Python implementation of Stata's egen command for pandas DataFrames. This package provides Stata-style data manipulation functions, making it easier for researchers to transition from Stata to Python while maintaining familiar syntax and functionality.

Quick Start

pip install pyegen

import pandas as pd
import numpy as np
import pyegen as egen

# Create sample data
df = pd.DataFrame({
    'group': ['A', 'A', 'B', 'B', 'C', 'C'],
    'var1': [1, np.nan, 3, 4, 5, 6],
    'var2': [np.nan, 2, 5, 6, 7, 8],
    'var3': [10, 11, 12, 13, 14, 15]
})

# Row-wise operations
df['first_nonmiss'] = egen.rowfirst(df, ['var1', 'var2', 'var3'])
df['row_median'] = egen.rowmedian(df, ['var1', 'var2', 'var3'])
df['missing_count'] = egen.rowmiss(df, ['var1', 'var2', 'var3'])

# Group-wise operations  
df['group_mean'] = egen.mean(df['var1'], by=df['group'])
df['group_median'] = egen.median(df['var1'], by=df['group'])
df['group_rank'] = egen.rank(df['var1'], method='min')

# Utility functions
df['has_value_1_or_2'] = egen.anymatch(df, ['var1', 'var2'], [1, 2])
df['concat_vars'] = egen.concat(df, ['group', 'var1'], punct='_')

Available Functions

PyEgen provides 45+ functions with 100% coverage of Stata's egen capabilities:

Row-wise Functions

• rowmean(), rowtotal(), rowmax(), rowmin(), rowsd()
• rowfirst(), rowlast(), rowmedian(), rowmiss(), rownonmiss(), rowpctile()

Statistical Functions

• rank(), count(), mean(), sum(), max(), min(), sd()
• median(), mode(), iqr(), kurt(), skew(), mad(), mdev()
• pc(), pctile(), std(), total()

Utility Functions

• tag(), group(), seq(), anycount(), anymatch(), anyvalue()
• concat(), cut(), diff(), ends(), fill()

🎯 Key Features

• Complete Stata Coverage: All 45 egen functions implemented
• Pandas Integration: Works seamlessly with pandas DataFrames
• Missing Value Handling: Consistent with Stata behavior
• Group Operations: Full support for by-group operations with by parameter
• Type Safety: Comprehensive input validation and error handling
• Performance: Optimized for large datasets

📚 Complete Function Reference

Row-wise Functions

Statistical Functions (with grouping support)

Utility Functions

💡 Migration Recommendation

For new projects, we recommend using the unified PyStataR package which provides a comprehensive suite of Stata-equivalent commands:

pip install py-stata-commands

from py_stata_commands import egen
df['rank_var'] = egen.rank(df['income'])

Why Consider PyStataR?

• Single installation for all Stata-equivalent commands (tabulate, egen, reghdfe, winsor2)
• Consistent API across all modules
• Enhanced documentation and examples
• Active development and long-term support

PyStataR Repository: https://github.com/brycewang-stanford/PyStataR

Documentation & Examples

For comprehensive examples and function documentation, see:

• Complete Function Reference
• Stata-to-PyEgen Mapping

📊 Function Coverage Status

• ✅ Row-wise functions: 11/11 (100%)
• ✅ Statistical functions: 17/17 (100%)
• ✅ Utility functions: 12/12 (100%)
• ✅ String functions: 2/2 (100%)
• ✅ Sequence functions: 2/2 (100%)

Total: 45/45 functions (100% coverage)

🧪 Testing

# Run tests
python -m pytest tests/

# Run specific test
python -m pytest tests/test_core.py

🔧 Project Status

PyEgen will continue to be maintained for existing users, but new feature development will primarily focus on PyStataR. This ensures:

• ✅ Bug fixes and compatibility updates for PyEgen
• ✅ Stable API for existing codebases
• 🚀 Enhanced features and new capabilities in PyStataR

Installation & Requirements

pip install pyegen

Requirements:

• Python 3.7+
• pandas >= 1.3.0
• numpy >= 1.20.0

🤝 Contributing

We welcome contributions! For major changes, please consider contributing to PyStataR for maximum impact.

🔗 Stata Documentation Reference

This implementation follows the official Stata documentation for egen:

• Stata 18 egen documentation

📄 License

This project is licensed under the MIT License - see the LICENSE file for details.

🔗 Related Projects

• PyStataR - Unified Stata-equivalent commands and R functions (recommended for new projects)
• StatsPAI - StatsPAI = Stats + Econometrics + ML + AI + LLMs