arabictest - اختبارات جذر الوحدة

# من GitHub
pip install git+https://github.com/merwanroudane/arabictest.git

# أو محلياً
cd arabictest
pip install -e .

import pandas as pd
import numpy as np
from arabictest import quick_test

# إنشاء بيانات تجريبية
np.random.seed(42)
data = pd.DataFrame({
    'GDP': np.random.randn(100).cumsum(),
    'Inflation': np.random.randn(100) * 0.5 + 2,
    'Exchange_Rate': np.abs(np.random.randn(100).cumsum()) + 100
})

# اختبار سريع مع عرض النتائج بالعربية
tester = quick_test(data, use_arabic=True)

Professional Unit Root Testing Package

Author: Dr. Merwan Roudane
Email: merwanroudane920@gmail.com
GitHub: https://github.com/merwanroudane/arabictest

Overview

A comprehensive Python package for performing unit root tests on economic time series with full Arabic language support and beautiful RTL text rendering.

Key Features

• Multiple Tests: ADF, KPSS, Phillips-Perron, DF-GLS
• Automatic Testing: at levels and first differences
• Log Transformation: automatic for positive variables
• Process Classification: distinguish between TS (Trend Stationary) and DS (Difference Stationary)
• Beautiful Tables: professional formatting with full Arabic support
• Export Results: to Excel and CSV formats
• Research-Ready: professional output for academic papers

Installation

# From GitHub
pip install git+https://github.com/merwanroudane/arabictest.git

# Or locally
cd arabictest
pip install -e .

Quick Start

import pandas as pd
import numpy as np
from arabictest import quick_test, UnitRootTester

# Create sample data
np.random.seed(42)
data = pd.DataFrame({
    'GDP': np.random.randn(100).cumsum(),
    'Inflation': np.random.randn(100) * 0.5 + 2,
    'Exchange_Rate': np.abs(np.random.randn(100).cumsum()) + 100
})

# Quick test with Arabic output
tester = quick_test(data, use_arabic=True)

# Or for English output
tester_en = quick_test(data, use_arabic=False)

Advanced Usage

Using the UnitRootTester Class

from arabictest import UnitRootTester

# Initialize tester
tester = UnitRootTester(data, use_arabic=True)

# Run all tests
tester.test_all_variables(
    tests=['ADF', 'KPSS', 'PP', 'DFGLS'],
    test_log=True,
    trend='ct'  # constant and trend
)

# Display results
tester.print_summary_table(tablefmt='fancy_grid')

# Get results as DataFrame
results_df = tester.get_summary_dataframe()

# Export to Excel
tester.export_to_excel('unit_root_results.xlsx')

# Export to CSV
tester.export_to_csv('unit_root_results.csv')

Testing a Single Variable

# Test specific variable
result = tester.test_single_variable(
    'GDP',
    tests=['ADF', 'KPSS', 'PP'],
    test_log=True,
    trend='c'  # constant only
)

# Display just this variable
tester.print_summary_table(var_name='GDP')

Batch Processing from Files

from arabictest import batch_test_from_excel, batch_test_from_csv

# Test all variables in Excel file
tester = batch_test_from_excel(
    'economic_data.xlsx',
    use_arabic=True,
    output_file='results.xlsx'
)

# Test all variables in CSV file
tester = batch_test_from_csv(
    'economic_data.csv',
    use_arabic=True,
    output_file='results.xlsx'
)

Compare Transformations

from arabictest import compare_transformations

# Compare results across transformations
comparison_df = compare_transformations(
    data['GDP'],
    var_name='GDP',
    use_arabic=True
)
print(comparison_df)

Test Options

Available Tests

• ADF: Augmented Dickey-Fuller test
• KPSS: Kwiatkowski-Phillips-Schmidt-Shin test
• PP: Phillips-Perron test
• DFGLS: Dickey-Fuller GLS test

Trend Specifications

• 'c': Constant only (default)
• 'ct': Constant and linear trend
• 'n': No constant or trend

Table Formats

Choose from various table formats:

• 'fancy_grid' (default): Beautiful box-drawing characters
• 'grid': Simple grid
• 'simple': Minimal formatting
• 'pipe': Markdown-style
• 'rst': reStructuredText
• 'html': HTML table

Understanding the Results

Process Types

The package automatically classifies series into:

• TS Process (Trend Stationary): Stationary around a deterministic trend
• DS Process (Difference Stationary): Non-stationary but becomes stationary after differencing
• Indeterminate: Requires further analysis

Interpretation

For ADF, PP, and DF-GLS tests:

• H₀: Unit root exists (non-stationary)
• H₁: No unit root (stationary)
• If p-value < 0.05 → Reject H₀ → Series is stationary

For KPSS test:

• H₀: Series is stationary
• H₁: Unit root present
• If p-value > 0.05 → Cannot reject H₀ → Series appears stationary

Arabic Output Examples

================================================================================
                      نتائج اختبار جذر الوحدة للمتغير: GDP                      
================================================================================
عدد المشاهدات: 100
نوع السلسلة: سلسلة مستقرة بالفروق (DS)
--------------------------------------------------------------------------------

المستوى
╒═══════════╤═══════════╤═══════════════════════╤═════════════╤═══════════════════════════════╤══════════════════════════════╕
│ الاختبار │ الإحصائية │ القيمة الاحتمالية    │ التأخيرات  │ القيم الحرجة                  │ النتيجة                     │
╞═══════════╪═══════════╪═══════════════════════╪═════════════╪═══════════════════════════════╪══════════════════════════════╡
│ ADF       │ -1.2345   │ 0.6543               │ 1           │ 1%: -3.46, 5%: -2.87, 10%: -2.57│ غير مستقر - يحتوي على جذر وحدة│
│ KPSS      │ 0.8765    │ 0.0123               │ 8           │ 1%: 0.74, 5%: 0.46, 10%: 0.35  │ غير مستقر - يحتوي على جذر وحدة│
│ PP        │ -1.3456   │ 0.6123               │ 5           │ 1%: -3.46, 5%: -2.87, 10%: -2.57│ غير مستقر - يحتوي على جذر وحدة│
╘═══════════╧═══════════╧═══════════════════════╧═════════════╧═══════════════════════════════╧══════════════════════════════╛

Dependencies

• numpy >= 1.20.0
• pandas >= 1.3.0
• scipy >= 1.7.0
• statsmodels >= 0.13.0
• arch >= 5.0.0
• tabulate >= 0.8.9
• prettytable >= 3.0.0
• arabic-reshaper >= 2.1.0
• python-bidi >= 0.4.2
• colorama >= 0.4.4

Examples

See the examples/ directory for comprehensive examples:

• basic_example.py: Basic usage
• advanced_example.py: Advanced features
• batch_processing.py: Processing multiple files
• custom_analysis.py: Custom analysis workflows

Testing

Run tests with pytest:

pytest tests/

Contributing

Contributions are welcome! Please feel free to submit a Pull Request.

License

MIT License - see LICENSE file for details.

Citation

If you use this package in your research, please cite:

@software{arabictest2024,
  author = {Roudane, Merwan},
  title = {arabictest: Professional Unit Root Testing with Arabic Support},
  year = {2024},
  url = {https://github.com/merwanroudane/arabictest}
}

Contact

Dr. Merwan Roudane
Email: merwanroudane920@gmail.com
GitHub: https://github.com/merwanroudane/arabictest