Thank you for the initializer list suggestion, Christoph. Nice to hear from
you.

Rasmus

On Mon, Apr 15, 2024 at 6:51 AM Christoph Hertzberg <[email protected]>
wrote:

> On 2024-04-05 01:01, Ian Bell wrote:
> > Is there a way to (in order of niceness):
> > 1. Get the comma init to work
>
> Using the comma-initializer results in a call-graph like:
>
>      operator,(operator,(...,
>          operator,(operator<<(MatrixXd,double),
>          ...), double), double)
>
> With a 45*45 call-depth this may stress your compiler more than it is
> capable. I would use the comma-initializer only if you need C++03
> compatibility or to compose larger matrices from (a few!) sub-matrices.
>
> > 2. Do a less-annoying workaround
>
> I suggest using much more readable initializer lists:
>
>      const Eigen::MatrixXd kernel_1{
>          {0.0, 0.01, 0.02, ..., 0.44},
>          {1.0, 1.01 ,1.02, ..., 1.44},
>          ...
>      };
>
>
> > 3. At least crash with some sort of useful message
>
> I guess that is a compiler issue. We can't do much from Eigen's side.
>
> Christoph
>
>
> --
>   Dr.-Ing. Christoph Hertzberg
>
>   DFKI GmbH
>   Robotics Innovation Center
>   Robert-Hooke-Straße 1
>   28359 Bremen, Germany
>
>   Tel.:     +49 421 178 45-4021 <+49%20421%20178454021>
>   Zentrale: +49 421 178 45-0
>   E-Mail:   [email protected]
>
>   Weitere Informationen: http://www.dfki.de/robotik
>
>    -------------------------------------------------------------
>    Deutsches Forschungszentrum für Künstliche Intelligenz GmbH
>    Trippstadter Straße 122, 67663 Kaiserslautern, Germany
>
>    Geschäftsführung:
>    Prof. Dr. Antonio Krüger (Vorsitzender)
>    Helmut Ditzer
>
>    Vorsitzender des Aufsichtsrats:
>    Dr. Ferri Abolhassan
>    Amtsgericht Kaiserslautern, HRB 2313
>    -------------------------------------------------------------
>
>
>
>

Reply via email to