Transf WD Euler transformation
CresitelloDittmar, Mark
mdittmar at cfa.harvard.edu
Fri Apr 3 17:46:03 CEST 2020
Hopefully, the gWCS folks can respond to this more fully.
It was added in support of their implementation, which I mapped to
EulerAngleRotation. However, looking at that spec
<https://docs.astropy.org/en/stable/api/astropy.modeling.rotations.EulerAngleRotation.html>,
it appears to support only 3 rotations.
Mark
On Fri, Apr 3, 2020 at 9:11 AM Laurent MICHEL <
laurent.michel at astro.unistra.fr> wrote:
> Hello,
>
> I do not practice Euler transformations daily, but according to my
> college souvenirs, that is by definition a 3D rotation with a well
> defined sequence of rotation axis.
> In my understanding, the Euler rotation is even not a particular case of
> the Transf rotation since the 3 Euler rotations are not about native
> axis as stated in the definition you give.
>
> > Defines a rotation operation in a 3-dimensional cartesian coordinate
> space, defined as a series of rotations about the native axes (x,y,z)
>
> If I'm right, the class should be renamed (RotationSequence ?). If not,
> my comment is no longer relevant.
>
> Laurent
>
> Le 02/04/2020 à 16:40, CresitelloDittmar, Mark a écrit :
> > Laurent,
> >
> > This came directly from the implementation group discussions back in Dec
> > 2018.
> > gWCS has use cases where more than 3 rotations are useful, for example
> > EulerRotation(angles=[12, 1, 2, 3], axes=‘xyzx’).
> > So we've taken this approach of having a sequence of EulerAngles which
> > specify the 'axis of rotation' (of the 3D space as either 'x', 'y', or
> > 'z') and angle.
> >
> > Mark
> >
> >
> > On Thu, Apr 2, 2020 at 8:31 AM Laurent MICHEL
> > <laurent.michel at astro.unistra.fr
> > <mailto:laurent.michel at astro.unistra.fr>> wrote:
> >
> > Dear,
> >
> > My comment #1: Why is the Euler Transformation class so complex?
> >
> > - It is always a 3D transformation, this there is no reason to setup
> a
> > 1-* aggregation with angles.
> >
> > - Euler angles (nutation, rotation, intrinsic rotation) are well
> > defined, there is no need for a specific datatype (EulerAngle). So we
> > could just have 3 class attributes.
> >
> > - Using a specific datatype where angles are identified by a the
> > name of
> > the rotation axis supposes we have a vocabulary for this, which is a
> > bit
> > complicated for this purpose with the risk a implementation errors.
> >
> > Take care of yourself.
> >
> > LM
> >
> > --
> > ---- Laurent MICHEL Tel (33 0) 3 68 85 24 37
> > Observatoire de Strasbourg Fax (33 0) 3 68 85 24 32
> > 11 Rue de l'Universite Mail
> > laurent.michel at astro.unistra.fr <mailto:
> laurent.michel at astro.unistra.fr>
> > 67000 Strasbourg (France) Web
> http://astro.u-strasbg.fr/~michel
> > ---
> >
>
> --
> ---- Laurent MICHEL Tel (33 0) 3 68 85 24 37
> Observatoire de Strasbourg Fax (33 0) 3 68 85 24 32
> 11 Rue de l'Universite Mail laurent.michel at astro.unistra.fr
> 67000 Strasbourg (France) Web http://astro.u-strasbg.fr/~michel
> ---
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.ivoa.net/pipermail/dm/attachments/20200403/7278155f/attachment.html>
More information about the dm
mailing list