Some convenience functions for Cosmology-related analysis.
Project description
- Date:
- March 18, 2021
docs |
|
---|---|
tests |
|
package |
|
conda-forge |
Introduction
This package contains some numba-jit-compiled functions that perform Quaternion operations and a convenient class Quaternion that provide convenient methods wrapping around those functions.
Quaternion behaves like a Numpy array containing quaternion, e.g. respect Numpy broadcast operations, but without really imitating a numpy.ndarray and implemented a dtype.
This design allows you to write any jit-compiled functions involving those provided jit-compiled functions, and then write your own class methods that calls those functions as a convenient interface (by class inheritance.)
If you do not care to use Quaternion in other jit-compiled functions you write, check out packages below instead.
Other Python quaternion projects
Other Python projects that implements quaternions and I knew and used are:
zonca/quaternionarray: written in pure Python using Numpy. Note that unusually they put the real part in the last column. lastcol_quat_to_canonical and canonical_quat_to_lastcol convert between those and the canonical ordering (where real part comes first.)
hpc4cmb/toast: toast.qarray is a reimplementation of the above quaternionarray package in C++ with the same interface, and following the same convention.
moble/quaternion: implement Quaternion as a Numpy dtype in C.
moble/quaternionic: implement Quaternion as a Numpy dtype using Numba. This package is inspired by my expectation of quaternionic—I expected I could use them in a Numba-jit-compiled function but it doesn’t.
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 numba_quaternion-0.1.0-py3-none-any.whl
Algorithm | Hash digest | |
---|---|---|
SHA256 | 5094d2103d1e30558ddd027f73a3b3b733e453fba4832f292c32e24038ae1f81 |
|
MD5 | a19bb702a0d33255eb5c5e7b50643956 |
|
BLAKE2b-256 | b55243d0a3ab41c1d171770b6a22bf3df22464eaefb01e5501c6a54ef4f60435 |