PauBadiaM / composition_stats

Python package for compositional data analysis

Composition Statistics

Python module for compositional data analysis. Install with pip:

pip install composition_stats

The following functions are provided:

closure	Performs closure to ensure that all elements add up to 1.
multiplicative_replacement	Replace all zeros with small non-zero alues
perturb	Performs the perturbation operation.
perturb_inv	Performs the inverse perturbation operation.
power	Performs the power operation.
inner	Calculates the Aitchson inner product.
clr	Performs centre log ratio transformation.
clr_inv	Performs inverse centre log ratio transformation.
ilr	Performs isometric log ratio transformation.
ilr_inv	Performs inverse isometric log ratio transform.
alr	Performs additive log ratio transformation.
alr_inv	Performs inverse additive log ratio transform.
center	Computes the geometric average of data.
centralize	Center data around its geometric average.
sbp_basis	Builds an orthogonal basis from a sequential binary partition (SBP).

Please see the documentation for details and a complete function reference.

This is a fork of the essential compositional data functions of the skbio.stats.composition module from scikit-bio. However, for reasons of performance, the functions in this module expect inputs where compositions already sum to unity, and do not call closure() on inputs internally.