Proxy classes for regression analysis

builds docs package wheel pyversions codecov coveralls scrutinizer

Overview

This package provides classes to easily model proxies for regression analysis. Originally intended to be used for (trace gas density) time series, but it might be useful in other cases too. The classes allow for a lag and finite proxy lifetime to be include when modelling the time series.

Development happens on github https://github.com/st-bender/pyregressproxy.

Install

Prerequisites

All dependencies will be automatically installed when using pip install or python setup.py, see below. However, to speed up the install or for use within a conda environment, it may be advantageous to install some of the important packages beforehand:

  • numpy at least version 1.13.0 for general numerics,
  • celerite at least version 0.3.0 for the “standard” model interface,
  • PyMC3, optional, for the PyMC3 model interface, and
  • PyMC4, optional, for the PyMC4 model interface.

numpy should probably be installed first because celerite needs it for setup. It may also be a good idea to install pybind11 beforehand because celerite uses its interface.

Depending on the setup, numpy and pybind11 can be installed via pip:

pip install numpy pybind11

or conda:

conda install numpy pybind11

regressproxy

Official releases are available as pip packages from the main package repository, to be found at https://pypi.org/project/regressproxy/, and which can be installed with:

$ pip install regressproxy

The latest development version of regressproxy can be installed with pip directly from github (see https://pip.pypa.io/en/stable/reference/pip_install/#vcs-support and https://pip.pypa.io/en/stable/reference/pip_install/#git):

$ pip install [-e] git+https://github.com/st-bender/pyregressproxy.git

The other option is to use a local clone:

$ git clone https://github.com/st-bender/pyregressproxy.git
$ cd pyregressproxy

and then using pip (optionally using -e, see https://pip.pypa.io/en/stable/reference/pip_install/#install-editable):

$ pip install [-e] .

or using setup.py:

$ python setup.py install

Usage

The module is imported as usual:

>>> import regressproxy

Basic class and method documentation is accessible via pydoc:

$ pydoc regressproxy

License

This python package is free software: you can redistribute it or modify it under the terms of the GNU General Public License as published by the Free Software Foundation, version 2 (GPLv2), see local copy or online version.