Rotation Matrices

Definition of rotation matrices

  • rotmat.R() … 3D rotation matrix for rotation about a coordinate axis

Conversion Routines

  • rotmat.convert() … Convert a rotation matrix to the corresponding quaternion
  • rotmat.sequence() … Calculation of Euler, Fick/aeronautic, Helmholtz angles
  • rotmat.seq2quat() … Calculation of quaternions from Euler, Fick/aeronautic, Helmholtz angles

Symbolic matrices

  • rotmat.R_s() … symbolix matrix for rotation about a coordinate axis

For example, you can e.g. generate a Fick-matrix, with

R_Fick = R_s(2, ‘theta’) * R_s(1, ‘phi’) * R_s(0, ‘psi’)

Details

Routines for working with rotation matrices

rotmat.R(axis=0, angle=90)[source]

Rotation matrix for rotation about a cardinal axis. The argument is entered in degree.

Parameters:
  • axis (skalar) – Axis of rotation, has to be 0, 1, or 2
  • alpha (float) – rotation angle [deg]
Returns:

R

Return type:

rotation matrix, for rotation about the specified axis

Examples

>>> rotmat.R(axis=0, alpha=45)
array([[ 1.        ,  0.        ,  0.        ],
       [ 0.        ,  0.70710678, -0.70710678],
       [ 0.        ,  0.70710678,  0.70710678]])
>>> rotmat.R(axis=0)
array([[ 1.        ,  0.        ,  0.        ],
       [ 0.        ,  0.        , -1.        ],
       [ 0.        ,  1.        ,  0.       ]])
>>> rotmat.R(1, 45)
array([[ 0.70710678,  0.        ,  0.70710678],
       [ 0.        ,  1.        ,  0.        ],
       [-0.70710678,  0.        ,  0.70710678]])
>>> rotmat.R(axis=2, alpha=45)
array([[ 0.70710678, -0.70710678,  0.        ],
       [ 0.70710678,  0.70710678,  0.        ],
       [ 0.        ,  0.        ,  1.        ]])
rotmat.R_s(axis=0, angle='alpha')[source]

Symbolic rotation matrix about the given axis, by an angle with the given name

Returns:R_symbolic
Return type:symbolic matrix for rotation about the given axis

Examples

>>> R_yaw = R_s(axis=2, angle='theta')
>>> R_nautical = R_s(2) * R_s(1) * R_s(0)
rotmat.convert(rMat, to='quat')[source]

Converts a rotation matrix to the corresponding quaternion. Assumes that R has the shape (3,3), or the matrix elements in columns

Parameters:
  • rMat (array, shape (3,3) or (N,9)) – single rotation matrix, or matrix with rotation-matrix elements.
  • to (string) – Currently, only ‘quat’ is supported
Returns:

outQuat – corresponding quaternion vector(s)

Return type:

array, shape (4,) or (N,4)

Notes

\[\begin{split}\vec q = 0.5*copysign\left( {\begin{array}{*{20}{c}} {\sqrt {1 + {R_{xx}} - {R_{yy}} - {R_{zz}}} ,}\\ {\sqrt {1 - {R_{xx}} + {R_{yy}} - {R_{zz}}} ,}\\ {\sqrt {1 - {R_{xx}} - {R_{yy}} + {R_{zz}}} ,} \end{array}\begin{array}{*{20}{c}} {{R_{zy}} - {R_{yz}}}\\ {{R_{xz}} - {R_{zx}}}\\ {{R_{yx}} - {R_{xy}}} \end{array}} \right)\end{split}\]

More info under http://en.wikipedia.org/wiki/Quaternion

Examples

>>> rotMat = array([[cos(alpha), -sin(alpha), 0],
>>>    [sin(alpha), cos(alpha), 0],
>>>    [0, 0, 1]])
>>> rotmat.convert(rotMat, 'quat')
array([[ 0.99500417,  0.        ,  0.        ,  0.09983342]])
rotmat.seq2quat(rot_angles, seq='nautical')[source]

This function takes a angles from sequenctial rotations and calculates the corresponding quaternions.

Parameters:
  • rot_angles (ndarray, nx3) – sequential rotation angles [deg]
  • seq (string) –

    Has to be one the following:

    • Euler … Rz * Rx * Rz
    • Fick … Rz * Ry * Rx
    • nautical … same as “Fick”
    • Helmholtz … Ry * Rz * Rx
Returns:

quats – corresponding quaternions

Return type:

ndarray, nx4

Examples

>>> skin.rotmat.seq2quat([90, 23.074, -90], seq='Euler')
array([[ 0.97979575,  0.        ,  0.2000007 ,  0.        ]])

Notes

The equations are longish, and can be found in 3D-Kinematics, 4.1.5 “Relation to Sequential Rotations”

rotmat.sequence(R, to='Euler')[source]

This function takes a rotation matrix, and calculates the corresponding angles for sequential rotations.

R_Euler = R3(gamma) * R1(beta) * R3(alpha)

Parameters:
  • R (ndarray, 3x3) – rotation matrix
  • to (string) –

    Has to be one the following:

    • Euler … Rz * Rx * Rz
    • Fick … Rz * Ry * Rx
    • nautical … same as “Fick”
    • Helmholtz … Ry * Rz * Rx
Returns:

  • gamma (third rotation (left-most matrix))
  • beta (second rotation)
  • alpha (first rotation(right-most matrix))

Notes

The following formulas are used:

Euler:

\[ \begin{align}\begin{aligned}\beta = -arcsin(\sqrt{ R_{xz}^2 + R_{yz}^2 }) * sign(R_{yz})\\\gamma = arcsin(\frac{R_{xz}}{sin\beta})\\\alpha = arcsin(\frac{R_{zx}}{sin\beta})\end{aligned}\end{align} \]

nautical / Fick:

\[ \begin{align}\begin{aligned} \theta = arctan(\frac{R_{yx}}{R_{xx}})\\\phi = arcsin(R_{zx})\\ \psi = arctan(\frac{R_{zy}}{R_{zz}})\end{aligned}\end{align} \]

Note that it is assumed that psi < pi !

Helmholtz:

\[ \begin{align}\begin{aligned}\theta = arcsin(R_{yx})\\\phi = -arcsin(\frac{R_{zx}}{cos\theta})\\\psi = -arcsin(\frac{R_{yz}}{cos\theta})\end{aligned}\end{align} \]

Note that it is assumed that psi < pi