Metadata-Version: 2.4
Name: extrastats
Version: 0.1.0
Summary: A Python package for advanced statistical analysis.
Project-URL: Homepage, https://github.com/jerradmgenson/extrastats
Project-URL: Issues, https://github.com/jerradmgenson/extrastats/issues
Author-email: Jerrad Michael Genson <jerradgenson@gmail.com>
License-Expression: MPL-2.0
License-File: LICENSE
Classifier: Operating System :: OS Independent
Classifier: Programming Language :: Python :: 3
Requires-Python: >=3.8
Requires-Dist: joblib==1.4.*
Requires-Dist: numpy==2.2.*
Requires-Dist: packaging==24.*
Requires-Dist: pandas==2.2.*
Requires-Dist: patsy==1.0.*
Requires-Dist: python-dateutil==2.9.0.post0
Requires-Dist: pytz==2024.2
Requires-Dist: scipy==1.15.*
Requires-Dist: six==1.17.*
Requires-Dist: statsmodels==0.14.*
Requires-Dist: tzdata==2025.1
Description-Content-Type: text/markdown

# extrastats

**Advanced statistical tools and routines for Python**

`extrastats` is a Python library that provides high-quality statistical methods to address gaps in mainstream libraries like NumPy, SciPy, and statsmodels. Designed for data scientists, statisticians, and researchers, `extrastats` includes robust, customizable, and performance-optimized routines.

Copyright 2022-2025 Jerrad Michael Genson

This library is licensed under the Mozilla Public License, v. 2.0.

---

## Features

### Robust Outlier Detection
- **Adjusted Boxplot**: A method that extends traditional boxplots using the medcouple statistic for skewness adjustment.

### Advanced Permutation Testing
- Flexible resampling strategies:
  - Pairings, samples, independent shuffles, and bootstrapping.
- Parallel computation with `joblib`.

### Confidence Interval Estimation
- Bootstrap-based confidence intervals for arbitrary statistics.
- Support for multiple confidence levels in a single call.

### Sample Size Estimation
- Monte Carlo simulations to determine required sample sizes for target confidence interval widths.

### Tail Weight Analysis
- Evaluate the tail weight of distributions using the L/RMC method.

### Mutual Information
- Compute mutual information for discrete variables.
- Optional normalization for interpretability.

### Additional Metrics
- Geometric Coefficient of Variation (GCV)
- Harmonic Variability (HVAR)
- Trimmed statistics and probability operations.

---
