| Re: R:Re: [eigen] GivensQR |
[ Thread Index |
Date Index
| More lists.tuxfamily.org/eigen Archives
]
- To: eigen@xxxxxxxxxxxxxxxxxxx
- Subject: Re: R:Re: [eigen] GivensQR
- From: Benoit Jacob <jacob.benoit.1@xxxxxxxxx>
- Date: Thu, 10 Sep 2009 07:55:06 -0400
- Dkim-signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=gmail.com; s=gamma; h=domainkey-signature:mime-version:received:in-reply-to:references :date:message-id:subject:from:to:content-type; bh=HQ9wxLOqchJujLRIqJfKxdYa7LxjAkrnfPMyrkahgdI=; b=kGWgxnIpMss9n8qHSyXjm1pkls1RpSkyKZyO3ooDAVQE8l3T1QDremlTAhHHEBxhVS srWFNhBiiQEaf1EvbUfqkw90Q/Pj0BWFHKqWXkXBD68e8cyAXIGEd44atoabcE+fqFXJ FDsa0o4K4bquN7FFvj4vfFkorS2d6JjaTFWdU=
- Domainkey-signature: a=rsa-sha1; c=nofws; d=gmail.com; s=gamma; h=mime-version:in-reply-to:references:date:message-id:subject:from:to :content-type; b=hILnbdx/Np7kpwLf8FvHAegNLFrmwCTCDir6YrRP8EIWS82jSM759FdUhOqkxDuIh/ D3ZW/qC8PHqZfbSEdK99cET1R6al549ZE5aVlY3C8SbDEAooefQg3U72JxWi7b+q5oop PQOHoeyNOs6sCcm2rUiYnuk3eBBe94nSlPFqE=
I see, I didn't think of that.
Let's first do the other changes and finalize the API and members,
then we'll see what to do about that.
Benoit
2009/9/10 <yo.eres@xxxxxxxxx>:
> I was also in doubt about that use of "mutable". The idea was that the
> computing of QR later than by the constructor is a const "action". For
> example, if one writes a function like this:
>
> template<typename MatrixType> void my_funct (const GivensQR<MatrixType>&
> qr);
>
> in which he uses a const version of a QR decomposition, then the function
> should maybe compute the decomposition. Besides, the matrixQ() and matrixR()
> methods should be const from the user's point of view, but these methods
> call the compute_xy(), so compute_xy() should also be const. This implies
> that all the properties should be mutable.
>
> On the other hand, if we decide that the stored matrices are not mutable,
> then the methods cannot be const...
>
> Andrea Arteaga