[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: Treelayout and CGS
From: |
Jaroslav Hajek |
Subject: |
Re: Treelayout and CGS |
Date: |
Fri, 14 Nov 2008 13:37:41 +0100 |
On Fri, Nov 14, 2008 at 12:59 PM, R! <address@hidden> wrote:
>
> Hi
>
> Thanks for your suggestion, I will look at the style issue and change the
> source code as it is mentioned in the documentation (sorry that I didn't do
> it before). But as I think that there will be also another suggestion I will
> wait some time with a new version.
>
> Radek Salac
>
Hi Radek,
this looks very nice. I have two comments to cgs:
1. Since calling `error' enters an exceptional state, no more function
code is executed (unless it's inside unwind_protect or try/catch
block). Therefore, you can write the error checks as
if (check1)
error ("error1")
elseif (check2)
error ("error2")
endif
... normal code ...
this saves you one indentation of the normal code.
2. What is the reason for supplying M0 & M1, if they are multiplied
anyway? I don't remember much from Krylov methods, but I guess perhaps
the purpose of specifying 2 preconditioners is so that their
multiplication can be avoided?
--
RNDr. Jaroslav Hajek
computing expert
Aeronautical Research and Test Institute (VZLU)
Prague, Czech Republic
url: www.highegg.matfyz.cz
- Treelayout and CGS, R!, 2008/11/11
- Re: Treelayout and CGS, David Bateman, 2008/11/14
- Re: Treelayout and CGS, David Bateman, 2008/11/14
- Re: Treelayout and CGS, R!, 2008/11/14
- Re: Treelayout and CGS,
Jaroslav Hajek <=
- Re: Treelayout and CGS, David Bateman, 2008/11/14
- Re: Treelayout and CGS, R!, 2008/11/19
- Re: Treelayout and CGS, David Bateman, 2008/11/19
- Re: Treelayout and CGS, R!, 2008/11/21
- Re: Treelayout and CGS, David Bateman, 2008/11/21
- Re: Treelayout and CGS, R!, 2008/11/25
- Re: Treelayout and CGS, David Bateman, 2008/11/25