statsmodels 0.14.4

Statistical computations and models for Python

Stars: 10055, Watchers: 10055, Forks: 2877, Open Issues: 2835

The statsmodels/statsmodels repo was created 13 years ago and the last code push was Yesterday.
The project is extremely popular with a mindblowing 10055 github stars!

How to Install statsmodels

You can install statsmodels using pip

pip install statsmodels

or add it to a project with poetry

poetry add statsmodels

Package Details

Author

None

License

BSD License

Homepage

https://www.statsmodels.org/

PyPi:

https://pypi.org/project/statsmodels/

Documentation:

https://www.statsmodels.org/stable/index.html

GitHub Repo:

https://github.com/statsmodels/statsmodels

Classifiers

• Office/Business/Financial
• Scientific/Engineering

Related Packages

Errors

A list of common statsmodels errors.

Code Examples

Here are some statsmodels code examples and snippets.

GitHub Issues

The statsmodels package has 2835 open issues on GitHub

• Unknown issue with - damped_trend=True
• OrderedResults.save: pickling not supported
• ENH: get_prediction inference for in-sample statistics, correlation to score_obs, ATE
• ENH: use scipy.stats.studentized_range in tukey hsd when available
• ENH: Treatment effect rebased
• ENH: heckman endogenous sample selection with non-gaussian errors
• ENH/REF/DOC improve hurdle and truncated count models
• GAM: help matching to R's mgcv package (interactions, random effects, categorical variables, etc)
• Structural VAR
• Distributed Estimation example for Seasonal-Trend decomposition using LOESS (STL)
• Theta forecasting model gives forecasts around 100 even if test data is in range of -5% to 5 %
• ENH: enhancements to hurdle count model, GLM or binary models as zero model
• ENH: hurdle count models - joint estimation and common parameters
• ENH: add zero-modified distribution
• ENH/Design: diagnostics based on linear predictor or tranformed distribution

See more issues on GitHub