On Tue, Oct 22, 2013 at 11:59 AM, F. B. <[email protected]> wrote:
>
> Other considerations: is SymPy able to match subtrees in the args tree? What
> I have understood from the docstrings, is that the matcher tries to match
> the whole expression tree, not a part of it.
>
> The tensor module is somewhat anomalous, because the way you write tensors
> is not the same as the expression tree, which contains additional
> information. I should be interesting to have a look at xAct into detail, I
> suppose they use something similar to assumptions or tools by Mathematica. I
> have not a very good knowledge of Mathematica language, though.
>
> In any case, subtree pattern matching could be an interesting matter, that's
> more graph-like than tree-like.
>
> I'm still convinced that the tensor module needs its own pattern matcher,
> the other way would be to relocate arg-tree internal data to the
> assumptions, or to some other global variable.

If you want to use a lot of mathematical knowledge about tensor
expressions that is not true for expressions in general, then a custom
pattern matcher will probably be more efficient and easier to write
than a generic one.

Aaron Meurer

>
> --
> You received this message because you are subscribed to the Google Groups
> "sympy" 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/sympy.
> For more options, visit https://groups.google.com/groups/opt_out.

-- 
You received this message because you are subscribed to the Google Groups 
"sympy" 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/sympy.
For more options, visit https://groups.google.com/groups/opt_out.

Reply via email to