Value space transformations

Graham Klyne gk at ninebynine.org
Mon Nov 10 11:43:38 EST 2003


I just found myself writing some code like this:

    processFoo processBar = toFoo . processBar . fromFoo
        where
            fromFoo foo = ...
            toFoo   bar = ...

and noticed some similarity in style to certain kinds of transformation 
used in linear algebra, such as for diagonalization:

    D = T M inv(T)

This got me to wondering if this is a common pattern in functional 
programming.  There seems to be some similarity, for example, with the way 
that Monads are sometimes used, where operations are "lifted" into (?) a Monad.

So my questions are:
(a) is there a common functional programming pattern that corresponds to 
vector space transformations so that a function defined over one space can 
be used in another, and
(b) if so, are there any not-too-heavy papers or articles discussing this 
pattern?

#g


------------
Graham Klyne
For email:
http://www.ninebynine.org/#Contact



More information about the Haskell-Cafe mailing list