[Date Index] [Thread Index] [Author Index]

Re: unable to FullSimplify

• To: mathgroup at smc.vnet.net
• Subject: [mg66010] Re: unable to FullSimplify
• From: "Vladimir" <vladimir347 at yahoo.com>
• Date: Wed, 26 Apr 2006 04:38:29 -0400 (EDT)
• References: <200604160545.BAA07958@smc.vnet.net> <200604181056.GAA14321@smc.vnet.net> <e24uhn$5s7$1@smc.vnet.net> <200604200914.FAA05331@smc.vnet.net> <e29qvc$m59$1@smc.vnet.net> <200604241001.GAA08924@smc.vnet.net> <81DDADBF-5461-4A2E-9F3C-6297C85B0C30@mimuw.edu.pl> <e2kqi2$7fd$1@smc.vnet.net>
• Sender: owner-wri-mathgroup at wolfram.com

Adam Strzebonski wrote:
> But this has a really huge complexity, and it still doesn't help

Hmm... Built-in FullSimplify performs exceptionally well
most of the time, i.e. it simplifies Expand[(a + b + c + d)^80],
which contains over a million leaves without any problem at all.
But it seems there is really small subset of problematic expressions.

So what's the reason for this problem?
Fundamental math, which prevents universal solution?
Or solution is possible but not desired due to performance reasons?

And the main question - how to fix it without the knowledge
of particular expression. If problematic pattern is rare and
well-defined, maybe it can be made to be "recognized" and
simplified as a "special case" without adding too much complexity?

And thanks for your detailed and insightful replies!

--
Vladimir

• References:
  • unable to FullSimplify
    • From: vladimir347@yahoo.com
  • Re: unable to FullSimplify
    • From: "Vladimir" <vladimir347@yahoo.com>
  • Re: unable to FullSimplify
    • From: "Vladimir" <vladimir347@yahoo.com>
  • Re: unable to FullSimplify
    • From: "Vladimir" <vladimir347@yahoo.com>