Perhaps my example was not clear enough. I start with a fully symbolic 
expression. Then, I try two things :
- ask a simplify on the fully symbolic expression, that works well,
- replace  some symbols in the initial expression and then simplify.

The first step shows that a simplified expression exists, and replacing by 
values  may not change this fact (it can lead to an even more simplified 
expresssion isn'nt it ?).

I do not understand why Sympy cannot simplify at all the initial expression 
when symbols are replaced by values, whereas we know that a simplified 
expresssion exists as Sympy itself found it from the fully symbolic version.



Le vendredi 31 juillet 2020 16:52:23 UTC+2, Oscar a écrit :
>
> On Sun, 26 Jul 2020 at 21:56, Mikhael Myara 
> <mikhae...@umontpellier.fr <javascript:>> wrote: 
> > 
> > And I obtain what follows. I am surprized that inserting values instead 
> of some symbols disables sympy simplification. In some more complex cases, 
> the computation times required for simplification are huge and lead to non 
> simplified expressions. 
>
> I think that inserting Float rather than Symbol or Rational can lead 
> to significant problems for some polynomial type simplifications such 
> as cancellation of factors. The long computation times might be the 
> result of converting Float to Rational somewhere which can sometimes 
> lead to very polynomials with high order or large coefficients. It's 
> hard to investigate that without self-contained code demonstrating the 
> problem though. 
>
> Do you have a reason to think that expression can be simplified much? 
> The main thing I notice is that it has repeating subexpressions so 
> something like cse might be useful: 
>
> https://docs.sympy.org/latest/modules/rewriting.html#common-subexpression-detection-and-collection
>  
>
> Oscar 
>

-- 
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 sympy+unsubscr...@googlegroups.com.
To view this discussion on the web visit 
https://groups.google.com/d/msgid/sympy/3a18afde-93ac-4236-a094-749a4e735c97o%40googlegroups.com.

Reply via email to