This goes a lot further than what I would need. I would not even need
the array indices to be symbolic expressions (although it certainly
would be nice to have that option). I would only like to substitute a
very large numpy array into a symbolic expression that is then
completely determined and should be evaluated (as efficiently as
possible). I understand of course that this is a bit specialized.
Nikolas
Am 27.01.2010 um 23:47 schrieb Aaron S. Meurer:
If I understand you correctly, it is related to this:
http://code.google.com/p/sympy/issues/detail?id=16
Symbols with indices is something that would be nice to have
implemented, but unfortunately isn't yet. So maybe the information
on that page will help you.
Aaron Meurer
On Jan 27, 2010, at 3:44 PM, Nikolas Tezak wrote:
Hi,
I am wondering if it were in principle possible to create subclass
of Symbol, that could handle array indices, so that at a later time
(after having performed symbolic calculations) I could just
substitute a numpy array in for that symbol.
e.g. (10 + a[:,0] *a[:,1]).subst(a, np.array([[...]])
Of course I could simply create symbols for individual array
elements, but I am working with a large number of coefficients
(~1000+) for which it would seem inefficient (at least to me) to
create individual symbols.
I am still trying to get a good feeling for the code and I will try
to implement it myself, but maybe some of you already know of some
reason why this would not work (efficiently).
Nikolas
--
You received this message because you are subscribed to the Google Groups
"sympy" group.
To post to this group, send email to [email protected].
To unsubscribe from this group, send email to
[email protected].
For more options, visit this group at
http://groups.google.com/group/sympy?hl=en.
