A module for calculation of PCA with the NIPALS algorithm
Project description
A module for calculation of PCA and PLS with the NIPALS algorithm. Based on the R packages nipals and pcaMethods as well as the statistical appendixes to “Introduction to Multi- and Megavariate Data Analysis using Projection Methods (PCA & PLS)” by Eriksson et. al. Tested to give same results as the above packages and Simca, with some rounding errors.
Free software: MIT license
Installation
pip install nipals
Documentation
Development
To run the all tests run:
tox
Note, to combine the coverage data from all the tox environments run:
Windows |
set PYTEST_ADDOPTS=--cov-append tox |
---|---|
Other |
PYTEST_ADDOPTS=--cov-append tox |
Changelog
0.5.2 (2019-06-04)
Added compatibility with Nipals objects saved from pre-0.5 versions
0.5.1 (2019-05-23)
Added checks for, and optional removal of, zero variance in variables
Added support for Python 3.7
(0.5.0 was never released due to failing CI tests)
0.4.3 (2018-04-24)
Fixed test that failed after last bug fix
0.4.2 (2018-04-24)
Fixed bug with selection of starting column for cross validation of PCA
0.4.1 (2018-04-09)
Fixed bug with cross validation of PCA
0.4.0 (2018-04-09)
Added cross validations
Added calculation of distance to model with plots
Added model overview plots
0.3.0 (2018-04-05)
Added R2X and R2Y to the PLS class
Made plot color selectable also for scoreplots without classes
0.2.0 (2018-03-29)
Added a PLS class
Improved plotting
Fixed some problems with missing/infinite values
0.1.0 (2018-03-14)
First release on PyPI.
Project details
Release history Release notifications | RSS feed
Download files
Download the file for your platform. If you're not sure which to choose, learn more about installing packages.
Source Distribution
Built Distribution
Hashes for nipals-0.5.2-py2.py3-none-any.whl
Algorithm | Hash digest | |
---|---|---|
SHA256 | 72a29393b80db473d74e6ace9bb6cf23da1cf574eb790ce1c02bdd4da628a7be |
|
MD5 | 317300fb6369a41134cf12c60906174b |
|
BLAKE2b-256 | dd46f53e21796c820fe06be53b6eb7575b91bc5dd4680507e50a5edb53b74fb9 |