The resulting symbolic expression must contain a different sort of 
information. For example

∂²f/[∂J(x1)∂J(x2)]

does not have a analogon, if you try to dervive it in terms of ordinary 
differentiation with subsequent substitution. Structurally, for the 
functional derivative, you must basically be able to differentiate a 
function of infinitely many arguments (thus on an infinitely dimensional 
space) with regard to one of those arguments. The ordinary partial 
derivative obviously does not capture this kind of operation.

I was contemplating that, perhaps, such an operation could be implemented 
on top of a function

*f: ℂ∞ ⊕ … ⊕ ℂ∞ → ℂ*
*
*
for f[J(x)] that would just be f: ℂ∞ → *ℂ. *But even if I'm able to express 
that function in SAGE, we'd still need some notion of a derivative on that. 
Perhaps there are other cathegorically similar structures which can be used 
to implement it, I don't know,


On Sunday, April 21, 2013 2:03:31 PM UTC+2, adsmith wrote:
>
> On Saturday, April 20, 2013 2:03:07 PM UTC-4, ManDay wrote:
>
>> I'd like to be able to derive in the following manner:
>>
>> x = var( "x" );
>> J = function( "g",x );
>> f = function( "f",g );
>>
>> # like so
>>
>> x1 = var( "x1" );
>>
>> f.diff( J( x1 ) );
>>
>> which is the functional f derivate with respect to the value of J at x1. 
>> Is this possible in a fashion at least similar to what I demonstrate above? 
>> I'm not understanding enough of the internal workings of symbolic 
>> expressions to modify "diff" so that it would do what I suggest, perhaps 
>> someone has a clever idea.
>>
>> Thanks.
>>
>
> How do you see this as being different than a substitution?
>  
>

-- 
You received this message because you are subscribed to the Google Groups 
"sage-devel" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to [email protected].
To post to this group, send email to [email protected].
Visit this group at http://groups.google.com/group/sage-devel?hl=en.
For more options, visit https://groups.google.com/groups/opt_out.


Reply via email to