Embeddings

[zmichels]

Hello,

The latest versions of MTEX (I am using 5.6.beta.1) includes support for defining "an embedding of the orientation space into an Eucledian space." At the bottom of the documentation page about embeddings (https://mtex-toolbox.github.io/OrientationEmbeddings.html)... it says...

Operations
The following operations are supported for embedding:

+, -, *, .*, ./
sum, mean
norm, normalize
dot
rotate, rotate_outer

However, for two embeddings defined from orientations, e1 and e2, if I try the operations *, .*, or ./, I get errors, such that these operations do not [yet] seem supported. I have also tried mtimes and other functions/operations, to no avail. Can anyone help me figure out how to use the * or .* operations on these?

Sidenotes:
I can apply those operations after converting to double (i.e., double(e1)), but then I have the trouble that:

1. I am not totally sure it is doing what I want, and
2. I am unsure how to get back to an embedding or orientation representation of the result (now a double).

Thanks for any help.

Cheers,
Zach

edit: for some reason, the comment window I am using does not publish the asterisk signs for the operations. I want to "multiply" them

[ralfHielscher]

Hi Zach,

until now the operation .* was indeed only implemented for products between a scalar and an embedding. In 28f3cbf I added support for embedding .* embedding as well as embedding ./ embedding . Would you like testing this?

I'm not so sure how to interpret embedding * embedding .

All the best,

Ralf.

[zmichels]

Hi Ralf, That’s great, I’ll check out that version. As far as the * operator, I guess I don’t know what I was expecting really either. I think I was just imagining using it similar to a rotation*rotation, but .* and ./ are great. Thank you! Zach

…

On Dec 31, 2020, at 12:40 AM, Ralf Hielscher ***@***.***> wrote: Hi Zach, until now the operation .* was indeed only implemented for products between a scalar and an embedding. In 28f3cbf <28f3cbf> I added support for embedding .* embedding as well as embedding ./ embedding . Would you like testing this? I'm not so sure how to interpret embedding * embedding . All the best, Ralf. — You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub <#658 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/ACC5AU2WS5RHTFK54AIA6XLSXQMFPANCNFSM4VOOI22Q>.