unistat Documentation (Version 0.3.0)#
Welcome to the documentation for unistat, a Python library to simplify
performing and reporting of medical, biostatistics, and social sciences
statistical analyses.
unistat is built on top of common Python data analysis & statistics
libraries, including pandas, SciPy,
and statsmodels. This library aims to implement best
practices for publication-quality statistical analysis, and to implement a
simple, straightforward API to to run these analyses, and view all data that is
pertinent for reporting in the context of academic publications.
Accordingly, unlike the statistics libraries on which unistat is built,
unistat is relatively opinionated: whereas parent libraries tend to
offer optionality in statistical methodologies, this library often selects
approaches that are generally accepted best practices for academic
publication, or at least are justifiable in such a setting. As a corollary,
unistat documentation aims to offer copious citations for its chosen
methods, so that a manuscript Methods section can appropriately justify
any methodological choices.
unistat’s raison d’être was to simplify statistics for medical (in
particular, surgical) clinical research; as such, to the extent that
accepted statistical methodologies in medical/surgical research differ from
other biostatistics or social sciences, the norms for clinical surgical
research will be prioritized. Nonetheless, where genuine methodological
optionality exists, some degree of choice is left to the user. At times,
this may not be easily accessible or obvious in the API, and in those cases,
users should access and choose non-default options only with informed
rationale for doing so.
This documentation covers unistat version 0.3.0, released
2026-03-06.
Getting Started#
Install unistat with:
pip install unistat
In future updates, more information will be available in the Getting Started guide.
Examples#
Coming in future updates to Examples.