Re: Bizarre semantic of Mat

[Ilya Zakharevich <nospam-abuse@ilyaz.org>]

On Wed, Jan 17, 2024 at 07:54:51AM +0100, Loïc Grenié wrote:
> > Today I (finally) looked through the documentation of Mat() — and it
> > seems to be very convenient for interactive usage!  On the other hand,

>      I confess I use the bizarre special-cases all the time...

Right: (as I said) it is very convenient if one knows the content of a vector…

>      Anyway, if you have a column C and want a 1-column matrix,
>   with C as only colon you can make Mat(vector(1,n,C)) (yes, I know,
>   this is anti-intuitive, long, non-convenient, not explicitly documented,
>   needed only because of bizarre semantics and uber-dirty).

Wow!!!  This is a very impressive ping-pong-with-APIs!

  It goes without saying that I would have rather preferred to have
  Mat(C, 1) to do this…

Thanks,
Ilya