Overview

This software provides a library of miscellaneous utilities / recipes for generic computations with Python and NumPy / SciPy. This code was originally included in FlyingCircus, but is now living in a separate package to avoid pulling larger dependencies to those users not in need of NumPy / SciPy functionalities. It is relatively easy to extend and users are encouraged to tweak with it.

Most of the code is used in a number of projects where it is tested against real-life scenarios.

All the code is tested against the examples in the documentation (using doctest).

The code has reached a reasonable level of maturity. However, until it gets a wider adoption, some of the library components may undergo some refactoring in the process of improving the code. Changes will appear in the CHANGELOG.rst. Please file a bug report if you detect an undocumented refactoring.

Releases information are available through NEWS.rst.

For a more comprehensive list of changes see CHANGELOG.rst (automatically generated from the version control system).

Features

The package (which requires both numpy and scipy – as well as FlyingCircus) – contains a number of numerical functions, typically working on or generating numpy.ndarray inputs, like:

• sgngeomspace(): generates geometrically / logarithmically spaced samples between signed start and stop endpoints.

• unsqueeze(): add singletons to the shape of an array to broadcast-match a given shape.

• subst(): conveniently substitute all occurrences of a value in an array.

etc.

These are meant to run both in Python 3

Installation

The recommended way of installing the software is through PyPI:

$ pip install flyingcircus_numeric

Alternatively, you can clone the source repository from GitHub:

$ git clone git@github.com:norok2/flyingcircus_numeric.git
$ cd flyingcircus_numeric
$ pip install -e .

For more details see also INSTALL.rst.

License

This work is licensed through the terms and conditions of the GPLv3+ See the accompanying LICENSE.rst for more details.

Acknowledgements

For a complete list of authors please see AUTHORS.rst.

People who have influenced this work are acknowledged in THANKS.rst.