rigid-body-motion 0.4.0

Python utilities for estimating and transforming rigid body motion.

rigid-body-motion

Python utilities for estimating and transforming rigid body motion.

Documentation: https://rigid-body-motion.readthedocs.io

Overview

This package provides a high-level interface for transforming arrays describing motion of rigid bodies between different coordinate systems and reference frames. The core of the reference frame handling is a fast re-implementation of ROS’s tf2 library using numpy and numpy-quaternion. The package also provides first-class support for xarray data types.

Installation

rigid-body-motion can be installed via pip:

$ pip install rigid-body-motion

or via conda:

$ conda install -c phausamann -c conda-forge rigid-body-motion

Examples

Transform numpy arrays between coordinate systems:

import numpy as np
import rigid_body_motion as rbm

arr_cart = np.ones((10, 2))
arr_polar = rbm.transform_coordinates(arr_cart, outof="cartesian", into="polar")

Transform numpy arrays across a tree of reference frames:

import numpy as np
import rigid_body_motion as rbm

rbm.register_frame("world")
rbm.register_frame("child", parent="world", translation=(1., 0., 0.))
rbm.register_frame("child2", parent="world", translation=(-1., 0., 0.))

arr_child = np.ones((10, 3))
arr_child2 = rbm.transform_points(arr_child, outof="child", into="child2")

Credits

This package was created with Cookiecutter and the audreyr/cookiecutter-pypackage project template.

History

0.4.0 (February 11th, 2021)

New features

• New lookup_linear_velocity and lookup_angular_velocity top-level methods.

• New render_tree top-level method for printing out a graphical representation of a reference frame tree.

• lookup_twist now accepts a mode parameter to specify the mode for angular velocity calculation.

Bug fixes & improvements

• Fixed a bug where estimated angular velocity was all NaN when orientation contained NaNs.

0.3.0 (December 8th, 2020)

New features

• Reference frames with timestamps now accept the discrete parameter, allowing for transformations to be fixed from their timestamp into the future.

• rbm accessor for DataArrays implementing qinterp and qinv methods.

• New best_fit_rotation and qinterp top-level methods.

Bug fixes & improvements

• Refactor of internal timestamp matching mechanism defining a clear priority for target timestamps. This can result in slight changes of timestamps and arrays returned by transformations but will generally produce more accurate results.

• Added mode and outlier_thresh arguments to estimate_angular_velocity.

• Fixed issues with iterative_closest_point.

0.2.0 (October 22nd, 2020)

New features

• New estimate_linear_velocity and estimate_angular_velocity top-level methods.

• New qmul top-level method for multiplying quaternions.

0.1.2 (October 7th, 2020)

Improvements

• Use SQUAD instead of linear interpolation for quaternions.

0.1.1 (September 17th, 2020)

Bug fixes

• Fix transformations failing for DataArrays with non-numeric coords.

0.1.0 (September 17th, 2020)

• First release