statsmodels/statsmodels

Statsmodels: statistical modeling and econometrics in Python

What it solves

Statsmodels provides a comprehensive suite of tools for statistical computation, estimation, and inference. It complements SciPy by offering a wider range of statistical models and tests, allowing users to perform complex data analysis and hypothesis testing without needing to implement these methods from scratch.

How it works

The library implements a vast array of statistical methodologies, including linear regression (OLS, WLS, GLS), generalized linear models (GLM), time series analysis (ARIMA, VARMAX, StateSpace modeling), and survival analysis. It also includes tools for nonparametric statistics, multivariate analysis (PCA, MANOVA), and data imputation techniques like MICE.

Who it’s for

It is designed for data scientists, statisticians, and researchers who need rigorous statistical modeling and inference tools in Python.

Highlights

  • Extensive Regression Models: Supports Ordinary Least Squares, Quantile regression, and Robust linear models.
  • Advanced Time Series Analysis: Includes a full StateSpace framework, Hidden Markov Models (HMM), and Vector Autoregressive models.
  • Diverse Statistical Tests: Provides a wide range of diagnostics, goodness-of-fit, and normality tests.
  • Multivariate Analysis: Features Principal Component Analysis with missing data and Factor Analysis.
  • Survival Analysis: Implements Cox proportional hazards regression and Kaplan-Meier estimation.

Related

  • Project
    maximtrp/scikit-posthocs

    A Python package providing a wide range of parametric and non-parametric post hoc tests for pairwise multiple comparisons in statistical data analysis.

  • Project
    geomstats/geomstats

    A Python package for computations, statistics, and machine learning on manifolds, enabling the analysis of non-Euclidean data like rotations and shapes.

  • Project
    pymc-devs/pymc

    PyMC is a Python package for Bayesian statistical modeling that uses MCMC and variational inference algorithms to estimate parameters in complex probabilistic models.

  • Project
    wilsonrljr/sysidentpy

    A Python library for system identification and time series forecasting using NARMAX models, offering tools for model structure selection and parameter estimation.

  • Project
    Nixtla/statsforecast

    A high-performance library for univariate time series forecasting that provides fast, scalable implementations of statistical and econometric models.