On Wednesday, July 13, 2011 12:29:39 PM UTC-7, Rob Beezer wrote:
>
> Thanks, Minh.  That's the way I see it as well, though as usual you 
> are more eloquent. 
>
> So this is really a two-part question.  Should  lift()  for free 
> module morphisms be deprecated? 
>
> If so, in favor of 
>
> (a) a simple renaming, as perhaps "pre_image()" 
>
> OR 
>
> (b) make the existing  inverse_image()  method more capable: 
> submodule in, submodule out  plus  element in, element out. 
>

I think of "preimage" and "inverse image" as synonyms, both meaning (as Nils 
says for inverse image) the set of elements in the domain mapping to the 
given element of, or subset of, the codomain.  As such, I think both terms 
have a precise mathematical meaning, and neither one will assign single 
elements to single elements: they assign a subset of the domain to a single 
element of the codomain.

For the given situation, I think that the word "lift" is ambiguous, but it 
may be the best choice.  Since it's ambiguous, it can handle the fact that 
f.lift(x) need not be mathematically well-defined; f.lift(x) is just some 
element from the preimage of x under f.  It's much better to use "lift" than 
to misuse "inverse_image".  Are there any other good candidates?  
"inverse_image_an_element" is accurate but clunky.  I don't think "pullback" 
would be appropriate.

I think extending the inverse_image() method is a good idea (just to 
singletons or to subsets?), and it should also be extended for morphisms of 
finitely generated modules over PIDs.

-- 
John

-- 
To post to this group, send an email to [email protected]
To unsubscribe from this group, send an email to 
[email protected]
For more options, visit this group at http://groups.google.com/group/sage-devel
URL: http://www.sagemath.org

Reply via email to