[Haskell] Re: [Haskell-cafe] ANNOUNCE: fad 1.0 -- Forward Automatic Differentiation library

Edward Kmett ekmett at gmail.com
Fri Apr 3 11:18:04 EDT 2009


A somewhat tricky concern is that that the extra functionality in question
depends on a bunch of primitive definitions that lie below this in the
package and the AD engine is used by a layer on top.

So moving it out would introduce a circular dependency back into the package
or require me to stratify into two packages.  When I looked into
partitioning the package for another reason I found that I couldn't do so
without introducing some orphan instances, so it'll probably be a tricky bit
of surgery to split out. That said, it's probably still worth doing.

I also agree that I should be somewhat more pedantic about the name. =)
-Edward Kmett

On Fri, Apr 3, 2009 at 10:49 AM, Barak A. Pearlmutter <barak at cs.nuim.ie>wrote:

> I feel silly, did not even notice that!  Thanks for the pointer.
>
> Would be sensible to merge the functionalities; will try to import
> functionality in Data.Ring.Module.AutomaticDifferentiation currently
> missing from Numeric.FAD.
>
> (One pedantic note: should really be named
> Data.Ring.Module.AutomaticDifferentiation.Forward, since it is doing
> forward-mode accumulation automatic differentiation; reverse is
> an adjoint kettle of fish.)
> --
> Barak A. Pearlmutter
>  Hamilton Institute & Dept Comp Sci, NUI Maynooth, Co. Kildare, Ireland
>   http://www.bcl.hamilton.ie/~barak/
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://www.haskell.org/pipermail/haskell/attachments/20090403/1b82d489/attachment.htm


More information about the Haskell mailing list