commons-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Phil Steitz <phil.ste...@gmail.com>
Subject Re: [math] SoC Commons-Math Status
Date Tue, 09 Aug 2005 19:28:59 GMT
The rule of thumb that we have tried to follow is that we never use
*only* a reference to NR and we avoid even looking at the code in NR.

On 8/9/05, Zhang <zxg_32@yahoo.com> wrote:
> Should we remove all the references to Numerical Recipes? I can use
> other books or URLs in the comments. 

Best to refer to other sources or at least include additional non-NR citations.

>Some of the materials are not
> included in NR, such as the divided difference representation of
> polynomials. 

Good, no problem with these as long as the implementations are
original and the source of the algorithms is not encumbered a la NR.

>Some methods however are pretty standard, like FFT. Of
> course we develop our own code but the underlying algorithm would
> be the same.

Best possible state - well-known, widely published algorithms
implemented with original code.  This is optimal both from the
standpoint of avoiding encumbered IP and maintainability and
reliability.

Phil

---------------------------------------------------------------------
To unsubscribe, e-mail: commons-dev-unsubscribe@jakarta.apache.org
For additional commands, e-mail: commons-dev-help@jakarta.apache.org


Mime
View raw message