[boost] Re: Math functions for Stastistics- feedback from TR2 proposal

From: Hubert Holin (Hubert.Holin_at_[hidden])
Date: 2004-10-22 10:33:44

Somewhere in the E.U., le 22/10/2004

   Bonjour

In article <cl62lb$3r3$1_at_[hidden]>,
 "Thorsten Ottosen" <nesotto_at_[hidden]> wrote:

> Hi Paul,
> 
> "Paul A Bristow" <pbristow_at_[hidden]> wrote in message
> news:E1CKELE-0006rT-8N_at_he204war.uk.vianw.net...
> | Sadly I was not able to attend The Redmond meeting to present proposals
> | myself,
> 
> Yeah, those who actually had implemented tr1 math functions was saying it was
> simply too hard to implement.

      You bet it's hard! *And* it takes time...

> | I find this response rather baffling considering the importance of these
> | functions for even the most basic statistics.
> 
> FWIW, I agree with that and mentioned that at the meeting.
> 
> | So I am now asking for your views on if it is worth producing a Boost
> | library version based on the existing Cephes implementation by Steven
> | Moshier who has kindly agreed to make it available under the Boost licence
> | terms.
> 
> I would be *very* interested in that since I need something to be able to move
> on with boost.stat

      As I have said before, yes, indeed, we do need a (boost) 
statistics library. If basing it on Cephes is the easiest path to get 
such a beast, then by all means let's use it! At any rate, this does not 
preclude modifications based on whatever input you can get (including of 
course what you can get here).

> | I am sure you will understand I would not want to put a lot of effort into
> | this, only to have it rejected by Boosters.
> |
> | Please let me have your views on if and how to progress this topic.

      Perhaps post a here a detailed overview of the task (what 
functions, their dependencies, a rough idea of how to implement them, 
already identified tough points...).

> I would of course vote strongly in favor of such a library. So please go
> forward with the idea about
> wrapping a implementation with a boost licence.
> 
> Once we have it in boost, I can start doing stuff in boost.stat. I hope this
> can be made into a standard proposal, which might then provide additional
> reason for the math functions to be in the standard.
> 
> best regards
> 
> Thorsten

      Agreed!

         Hubert Holin