Status: Accepted
Owner: ness...@googlemail.com
Labels: Type-Defect Priority-Medium Series
New issue 2641 by ness...@googlemail.com: limit() heuristics are wrong
http://code.google.com/p/sympy/issues/detail?id=2641
In [10]: limit(log(x)/z - log(2*x)/z, x, 0)
Out[10]:
2⋅zoo
─
z
In [11]: gru
Updates:
Labels: NeedsReview smichr
Comment #13 on issue 2641 by smi...@gmail.com: limit() heuristics are wrong
http://code.google.com/p/sympy/issues/detail?id=2641
With https://github.com/sympy/sympy/pull/1307 this gives
limit(log(x)/z - log(2*x)/z, x, 0)
-log(2)/z
--
You received t
Updates:
Status: Fixed
Comment #14 on issue 2641 by smi...@gmail.com: limit() heuristics are wrong
http://code.google.com/p/sympy/issues/detail?id=2641
(No comment was entered for this change.)
--
You received this message because you are subscribed to the Google Groups
"sympy-issues"
Issue 2641: limit() heuristics are wrong
http://code.google.com/p/sympy/issues/detail?id=2641
This issue is now blocking issue 2642.
See http://code.google.com/p/sympy/issues/detail?id=2642
--
You received this message because you are listed in the owner
or CC fields of this issue, or because yo
Comment #2 on issue 2641 by asmeurer: limit() heuristics are wrong
http://code.google.com/p/sympy/issues/detail?id=2641
Or change limit to not distribute over an Add, which is not always valid
(hence we get these nan problems all the time, because the case where it's
not valid is oo - oo ==
Comment #3 on issue 2641 by asmeurer: limit() heuristics are wrong
http://code.google.com/p/sympy/issues/detail?id=2641
For example, this works just fine if you combine it into a single term:
In [6]: limit((log(x) - log(2*x))/z, x, 0)
Out[6]:
-log(2)
───
z
--
You received this message b
Comment #4 on issue 2641 by ness...@googlemail.com: limit() heuristics are
wrong
http://code.google.com/p/sympy/issues/detail?id=2641
I think the idea [I did not write the code, of course] is that limit() does
all sorts of stuff which is sometimes invalid (for speed), and recognises
when
Updates:
Cc: ondrej.c...@gmail.com smi...@gmail.com
Comment #5 on issue 2641 by asmeurer: limit() heuristics are wrong
http://code.google.com/p/sympy/issues/detail?id=2641
Well, Ondrej wrote most of it, and I think Chris played around in there a
bit too. Perhaps they can share some in
Comment #6 on issue 2641 by smi...@gmail.com: limit() heuristics are wrong
http://code.google.com/p/sympy/issues/detail?id=2641
My initial feeling is that limit should do preprocecssing needed so it can
accurately decide whether guntz is needed. You have identified a case where
it makes sens
Comment #7 on issue 2641 by asmeurer: limit() heuristics are wrong
http://code.google.com/p/sympy/issues/detail?id=2641
It seems to me that you can't rely on an Add being combinable into
something else via together() or factor() or something like that.
By the way, is there a reason that grun
Comment #8 on issue 2641 by asmeurer: limit() heuristics are wrong
http://code.google.com/p/sympy/issues/detail?id=2641
For example, you can't factor out this expression into a non-Add:
In [25]: limit(log(x)*z - log(2*x)*y, x, 0)
Out[25]: zoo⋅y + zoo⋅z
In [26]: gruntz(log(x)*z - log(2*x)*y, x,
Updates:
Labels: WrongResult
Comment #9 on issue 2641 by asmeurer: limit() heuristics are wrong
http://code.google.com/p/sympy/issues/detail?id=2641
(No comment was entered for this change.)
--
You received this message because you are subscribed to the Google Groups
"sympy-issues" gro
Comment #10 on issue 2641 by asmeurer: limit() heuristics are wrong
http://code.google.com/p/sympy/issues/detail?id=2641
By the way, 2*zoo/z is also wrong. It should be auto-reduced to zoo/z.
But this is a separate issue.
--
You received this message because you are subscribed to the Googl
Comment #11 on issue 2641 by asmeurer: limit() heuristics are wrong
http://code.google.com/p/sympy/issues/detail?id=2641
On the other hand, in issue 2670, expanding it (so that it can distribute
across the Add) gives the right answer, but the factored result is wrong
(to be sure, this is als
14 matches
Mail list logo
