Re: [eigen] cwise ops nomenclature |
[ Thread Index |
Date Index
| More lists.tuxfamily.org/eigen Archives
]
- To: eigen@xxxxxxxxxxxxxxxxxxx
- Subject: Re: [eigen] cwise ops nomenclature
- From: Benoit Jacob <jacob.benoit.1@xxxxxxxxx>
- Date: Tue, 25 May 2010 03:41:50 -0400
- Dkim-signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=gmail.com; s=gamma; h=domainkey-signature:mime-version:received:received:in-reply-to :references:date:message-id:subject:from:to:content-type :content-transfer-encoding; bh=0sfAO71WzxowgGD4IOAbE/sKHBI7TKASCNgXYW1DLGU=; b=UMyPX5HVj35gOQslaUp+EETjD4kpkto1QMICaVY3XMcWc7CbxTIDmh8q8WGgPU+nd4 w2ZPfMtHHaTuTKzB1SKLts8FnrymZXbgt4G0uGzQCJeFFYKx2lS7OQY9oKGcAB0aT4YF apjn7ELEcc2snGw9s7ImKYrxV9Spma+JGgI+4=
- 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:content-transfer-encoding; b=lDLCiMCSRm+WpMCP1qJsWOpcHekY+xksjGWF3OVUxsw030E8pSxVJQLn5HLbZ7BIBB hDRbWpAbxL7C7x2kC45nVlsMSDV29BWNlHs4lC8VfssuxrZ4oTAIyTMkTZSG5TtJQ5Bh pVBZAxSGWnLEmmABd8E6HlPb1E+WoZ5KgS6L4=
Sorry, I'm sorry if, not being a native English speaker, I didn't pick
the best word, but it's too late now to make such a sweeping
nomenclature change, unless a swarm of native/good English speakers
here say it's really absolutely necessary...
Benoit
2010/5/24 Manoj Rajagopalan <rmanoj@xxxxxxxxx>:
> Hi eigen developers,
>
> This suggestion might be a little presumptuous but here it is anyway: I
> think cwise<Op>() member functions could be better understood and remember if
> renamed to ewise<Op>() or elemwise<Op>() or elementwise<Op>(). Here is why I
> think so:
>
> 1. I keep getting dragged into thinking that cwise<Op>() is column-wise as
> opposed to coefficient-wise. I think this is because I was introduced to
> cwise... and colwise() on the same day when I read the Eigen tutorial page.
> This might be the case with others who get introduced to Eigen.
>
> 2. The contents of matrices and vectors are usually referred to as elements
> more than coefficients. A "coefficient" is an entity that usually accompanies
> something else (as a multiplicative factor). In this sense, matrices (which
> are representations of linear operators in a certain basis) often are the
> coefficients themselves in a polynomial or multidimensional ODE (resulting
> from discretization of PDEs). I do suppose the elements of a matrix can be
> called coefficients because they do multiply components of vectors to whcih
> they are applied ...
>
> cheers!
> Manoj
>
>
>