Re: Multigrid convergence

And to comment also on some other points:

Han de Bruijn <Han.deBruijn@xxxxxxxxxxxxxx> writes:

Non-pde problems ? Can you mention examples of these, please ?

Google for "AMG non-pde"

Well, I'd like to compare this with the existence of the Euler number e.
Numerical analists have found a series 1 + 1/2! + 1/3! + .. + 1/n! and
they have found a limit (1 + 1/n)^n and they have .. . But this author
has found that all of this numerical stuff boils down to a number e with
the property that ln(e) = 1 . What I mean to say is that my idealized MG
in one dimension has _essentially_ all the ingredients of "real" MG. And
I would hesitate to call this "narrow minded".

If it implies that "real multigrid" works only for M-matrices, then it is
narrow-minded.

[...] In short:

Nothing is more practical than a good theory.

Yes, if it is good.

Nicolas Neuss
.

Relevant Pages

  • Re: Set inclusion and membership
    ... Han de Bruijn writes: ... > Look who is saying this! ... Just go to Google and search for the phrase ...
    (sci.math)
  • Re: Cantorian pseudomathematics
    ... But all of his arguing simply boils down to the following: ... Han de Bruijn ... Prev by Date: ...
    (sci.math)
  • Re: New Newsgroups
    ... In message, Kate ... Obviously it boils down to the question of who can see the future and ... I will transfer the site from Google to a private one if I have ...
    (uk.rec.sheds)
  • Re: what is Random ?
    ... Alexander S.F.X. van Dusen wrote: ... I Google a lot and collected some information on that topic, ... Han de Bruijn ...
    (sci.math)
  • Re: Multigrid convergence
    ... Nicolas Neuss wrote: ... Han de Bruijn writes: ... has found that all of this numerical stuff boils down to a number e with ...
    (sci.math.num-analysis)