Re: [eigen] cross product with complex vectors |

[ Thread Index |
Date Index
| More lists.tuxfamily.org/eigen Archives
]

Thanks for the quick answer. That helps for "mixed" vectors, if one does
not have an imaginary part.
Orell
On 07/16/2018 01:53 PM, Christoph Hertzberg wrote:
> Here is a related bugzilla entry:
>
> http://eigen.tuxfamily.org/bz/show_bug.cgi?id=944
>
> Short story: We can't just change our API here, but you can use this as
> a workaround:
>
> a.cross(b.conjugate());
>
>
> Christoph
>
>
> On 2018-07-16 13:35, Orell Garten wrote:
>> Hi
>>
>> I was wondering whether the implementation of the cross product for
>> complex vectors [1] is correct.
>>
>> Assume I have two vectors of type Vector3cd which hold a=(0,0,1) and
>> b=(0,1+i,0). I do a.cross(b) and would expect to get (-1-i,0,0) but get
>> (-1+i,0,0) as a result.
>>
>> I think the formula in [1] should be (a + ib) x (c x id) = (a x c - b x
>> d) + i (a x d + b x c) but maybe I am wrong.
>>
>> It would be great if somebody could provide some sanity check or the
>> reasoning behind [1].
>>
>> Thanks for helping,
>> Orell
>>
>> [1]
>> http://eigen.tuxfamily.org/dox/group__Geometry__Module.html#ga0024b44eca99cb7135887c2aaf319d28
>>
>>
>>
>>
>>
>>
>